Arrested apoptosis of nurse cells during Hydra oogenesis and embryogenesis

Available online at www.sciencedirect.com R

Developmental Biology 260 (2003) 191–206

www.elsevier.com/locate/ydbio

Arrested apoptosis of nurse cells during Hydra oogenesis and embryogenesis Ulrich Technau,a,* Michael A. Miller,b,1 Diane Bridge,b,2 and Robert E. Steeleb b

a Molecular Cell Biology, Darmstadt University of Technology, Schnittspahnstrassc 10, 64287 Darmstadt, Germany Department of Biological Chemistry and the Developmental Biology Center, University of California, Irvine, CA 92697-1700, USA

Received for publication 22 May 2002, revised 18 April 2003, accepted 21 April 2003

Abstract During Hydra oogenesis, an aggregate of germ cells differentiates into one oocyte and thousands of nurse cells. Nurse cells display a number of features typical of apoptotic cells and are phagocytosed by the growing oocyte. Yet, these cells remain unchanged in morphology and number until hatching of the polyp, which can occur up to 12 months later. Treatments with caspase inhibitors can block oocyte development during an early phase of oogenesis, but not after nurse cell phagocytosis has taken place, indicating that initiation of nurse cell apoptosis is essential for oocyte development. The genomic DNA of the phagocytosed nurse cells in the oocyte and embryo shows large-scale fragmentation into 8- to 15-kb pieces, but there is virtually none of the internucleosomal degradation typically seen in apoptotic cells. The arrested nurse cells exhibit high levels of peroxidase activity and are prevented from entering the lysosomal pathway. After hatching of the polyp, apoptosis is resumed and the nurse cells are degraded within 3 days. During this final stage, nurse cells become TUNEL-positive and enter secondary lysosomes in a strongly degraded state. Our results suggest that nurse cell apoptosis consists of caspase-dependent and caspase-independent phases. The independent phase can be arrested at an advanced stage for several months, only to resume after the primary polyp hatches. © 2003 Elsevier Science (USA). All rights reserved. Keywords: Apoptosis; Hydra; Oogenesis; Embryogenesis; Nurse cell; Cnidaria; Peroxidase; Caspase; Lysosome

Introduction Oocyte growth in animals frequently depends on nutrition by surrounding nurse cells. In many cases these nurse cells are derived from the germline, as is the oocyte. In Drosophila, for example, a cyst forms around 16 germ cells that remain interconnected by cytoplasmic bridges, one of which is determined to form the oocyte, while the remaining 15 cells differentiate into nurse cells and provide nutritive substances to the oocyte (King, 1970). In the course of oocyte development, the nurse cells undergo apoptosis and * Corresponding author. Molekulare Zellibiologie, TU Darmstadt, Schnittspahnstr. 10, 64287 Darmstadt, Germany, Fax: ⫹49-6151-166077. E-mail address: [email protected] (U. Technau). 1 Present address: Department of Cell and Developmental Biology, Vanderbilt University School of Medicine, Nashville, TN 37232; USA. 2 Present address: Department of Biology, Elizabethtown College, Elizabethtown, PA 17022-2298, USA.

are finally resorbed by the oocyte (Spradling, 1993; Cavaliere et al., 1998; McCall and Steller, 1998). Comparative analyses have shown that apoptosis of germline-derived nurse cells is widespread in the animal kingdom and often a necessary requirement for the proper development of the oocyte (for review see Matova and Cooley, 2001). In Drosophila, the caspase-3-deficient mutant dcp-1 does not form fertile oocytes (McCall and Steller, 1998). In C. elegans, half of the female germ cells undergo apoptosis and are engulfed by the surrounding sheath cells (Gumienny et al., 1999). The programmed cell death of these cells appears to be required for the maintenance of germline homeostasis because full-size oocytes do not form in the apoptosis-deficient mutants ced-3 and ced-4 (Matova and Cooley, 2001; Gumienny et al., 1999). Yet, both in Drosophila and C. elegans, the pathway for germ cell death seems to differ from apoptosis in other cell lineages, as some of the crucial regulators involved in somatic cell

0012-1606/03/$ – see front matter © 2003 Elsevier Science (USA). All rights reserved. doi:10.1016/S0012-1606(03)00241-0

192

U. Technau et al. / Developmental Biology 260 (2003) 191–206

Fig. 1. Acridine orange (AO) staining during oogenesis and embryogenesis. (A–D) Oogenesis stages. (A) Oogenesis stage 2 showing a faint background staining due to the high cell density in the egg field of aggregated germ cells. At this stage the oocyte (arrow) becomes determined and nurse cells start differentiating. (B) At stage 3, when the oocyte starts phagocytosing the nurse cells, a small number of dispersed AO-positive nurse cells that have been phagocytosed by the oocyte appear in the egg field. (C) At stage 5, the oocyte has phagocytosed most of the surrounding nurse cells, which are all brightly stained by AO. The arrows indicate the oocyte pseudopods which are being retracted at this stage. (D) Mature oocyte (stage 7) with thousands of incorporated AO-positive nurse cells. (E, F) Embryogenesis. (E) Eight-cell cleavage stage embryo (F) gastrula stage embryo (G) polyp immediately after hatching (8 weeks after fertilization) with a nearly equal density of AO-positive nurse cells as in oocytes and embryos. Note the orange fluorescence of AO-positive nurse cells in the tentacles (arrow), indicating that the nurse cells are being degraded; (H) 2-day-old polyp showing a significantly reduced number of nurse cells relative to a newly hatched polyp. Note the gradient of nurse cell disappearance from the tentacles (arrow) towards the aboral end of the polyp. (I) Close-up of a 2-day embryo showing red (arrowheads) and green fluorescent AO-positive nurse cells (arrows). Bars, 200 (A–H), 150 (G, H), and 200 ␮m (I).

apoptosis (i.e., EGL-1 in C. elegans and H99 genes in Drosophila) do not play a role in apoptosis of the germ cells (Foley and Cooley, 1998; Gumienny et al., 1999). More-

over, apoptosis of nurse cells in Drosophila involves highly regulated intracellular rearrangements of the actin cytoskeleton, which are important for the transfer of cytoplasmic

U. Technau et al. / Developmental Biology 260 (2003) 191–206

193

components into the oocyte (McCall and Steller, 1998; Mahajan-Miklos and Cooley, 1994a, 1994b; de Cuevas and Spradling, 1998; Matova et al., 1999). Such cytoskeletal rearrangements have not been reported in association with somatic cell apoptosis. In Hydra, a representative of the early-diverging metazoan phylum Cnidaria, thousands of germline-restricted interstitial stem cells aggregate to form the oocyte (Kleinenberg, 1872; Nishimiya-Fujisawa and Sugiyama, 1995; Littlefield, 1991). Although many of these cells are competent to form an oocyte, under normal conditions only one located in the middle of the aggregation is selected (Miller et al., 2000). The cells which are not selected differentiate into nurse cells and are phagocytosed by the growing oocyte. During and following their phagocytosis, the nurse cells begin a process in which they acquire a number of features of apoptotic cells (Honegger, 1989; Miller et al., 2000): the nuclear membrane becomes irregular, but retains its integrity; the chromatin becomes condensed and concentrates on one side of the cell; the cell rounds up and forms strong interdigitations at the contact sites with adjacent nurse cells. Subsequently the volume of the cell decreases, presumably due to loss of cytoplasmic fragments. In a preceding study, we have shown that initiation of the nurse cell apoptotic program is regulated by the oocyte (Miller et al., 2000). Apoptosis has been reported from Hydra and other Hydrozoa and attributed to growth regulation and morphogenesis (Bosch and David, 1984; Cikala et al., 1999; Seipp et al., 2001). Recently, apoptosis has also been shown to occur during spermatogenesis in Hydra (Kuznetsov et al., 2001). Here, we have examined the process of nurse cell apoptosis during oogenesis on the cellular and molecular level. We show that the early but not the late stages of oogenesis are caspase-dependent. After this caspase-dependent phase, the nurse cells are phagocytosed by the oocyte and arrest at an advanced stage of apoptosis. Arrested nurse cells are characterized by the presence of high levels of peroxidase activity and genomic DNA that is cleaved into large fragments. The internucleosomal cleavage pattern typical of apoptotic cells and entry into the lysosomal pathway does not occur until embryogenesis is completed.

Materials and methods Hydra culture, and induction and staging of oogenesis

Fig. 2. Phagocytosing early oocyte and morphology of incorporated nurse cells. (A) Macerated early stage 3 oocyte with the germinal vesicle in the center (arrow). (B) The same oocyte stained with DAPI, showing the nuclei of the phagocytosed nurse cells. (C) Phagocytosed nurse cell in a mature oocyte double stained with AO and DAPI showing that AO stains both nuclear (DNA; stained blue) and cytoplasmic (RNA; stained green) material. Note the peripheral location of the DNA. The dotted line indicates the periphery of the nurse cell. Bars 100 (A,B) and 20 ␮m (C), respectively.

Animals were maintained as described previously (Technau and Holstein, 1996). All experiments were carried out using the AEP female strain of Hydra vulgaris (Martin et al., 1997). Fertilization of eggs was done with males of the PA2 strain. Stock cultures of AEP were fed with Artemia nauplii five times a week, which kept them in a nonsexual state. To induce gametogenesis, animals were starved for one week and thereafter fed two to three times a week for several weeks (modified after Hobmayer et al., 2001). The

194

U. Technau et al. / Developmental Biology 260 (2003) 191–206

first stages of oogenesis became visible in such cultures about 10 –14 days after the initiation of starvation. Staging of oocyte development was done according to Miller et al. (2000). Acridine orange and DAPI staining Live animals were incubated in 10 ␮g/ml acridine orange (AO) in Hydra medium for 3 min in the dark, followed by five washes in Hydra medium (Miller et al., 2000). The animals were then relaxed by incubation in 0.03% heptanole or 0.5% urethane in Hydra medium for the time of examination under the microscope. Stained animals were mounted on glass slides and photographed immediately under epifluorescent illumination. DAPI staining was carried out on unfixed material (in the case of double staining of nurse cells with AO) or on dissected egg fields which were macerated on glass slides and then fixed with formaldehyde. Maceration was done by incubating in maceration solution (acetic acid/glycerol/water; 1:1:26; David, 1973) for 30 min and fixed on gelatine-coated glass slides with 4% formaldehyde.

terstained with 0.05 ␮g/ml DAPI and the macerates were mounted in glycerol/PBS (9/1). Hydroxyurea treatment Animals were treated with 10 mM hydroxyurea in Hydra medium for 8 h, and then washed several times and analyzed after 12–18 h. Antibody staining The monoclonal antibody H22.58 was generated in the lab of Charles N. David (Munich) and detects secondary lysosomes (Technau, David, and Holstein, in preparation). Animals were relaxed for 1 min in 2% urethane and fixed with Lavdovsky’s fixative (4% formaldehyde, 10% acetic acid, 50% ethanol) overnight. Animals were washed three times with PBS and incubated overnight in undiluted mAb H22.58. After three washes in PBS the animals were incubated for 2 h in 1:400 anti-mouse IgG and IgM Alexa 488 coupled secondary antibody (Molecular Probes). After three washes in PBS the specimen were mounted in glycerol/ PBS (9:1). Double staining in macerates with TUNEL was performed after the TUNEL assay.

TUNEL assay Caspase inhibitor treatment The TUNEL assay (Gavrieli et al., 1992) was adapted from Hensey and Gautier (1998). Animals were relaxed in 2% urethane for 1 min, fixed in MEMFA (100 mM MOPS, pH 7.4, 2 mM EGTA, 1 mM MgSO4, 3.7% formaldehyde) for 1 h and washed several times in methanol. The methanol was replaced stepwise with PBS, and then the animals were washed twice for 15 min in PBS/0.2% Tween and once in PBS. The animals were then equilibrated in terminal deoxytransferase (TdT) buffer for 1 h. End-labeling was carried out overnight at room temperature in 100 ␮l TdT buffer containing 0.5 ␮M digoxygenin-dUTP and 150 U/ml TdT. The reaction was stopped by addition of PBS/EDTA at 65°C (2⫻1 h), followed by four washes in PBS. The polyps were then washed in PBS/0.1% Triton X-100/2% BSA for 15 min and incubated in blocking serum (Roche) for 1 h. Then the specimens were incubated in a 1:2000 dilution of anti-digoxigenin antibody coupled to alkaline phosphatase (Roche) overnight at 4°C. Over the course of the following day, the samples were washed with a number of changes of PBS/Triton/BSA. For detection, the animals were equilibrated 2 ⫻ 5 min in alkaline phosphatase buffer (100 mM Tris, pH 9.5, 50 mM MgCl2, 100 mM NaCl, 0.1% Tween 20, 1 mM Levamisol) and then the substrates NBT and BCIP were added. The color reaction was stopped with 95% ethanol and the animals were mounted in glycerin/PBS (9:1). In macerates (David, 1973) the TUNEL assay was performed essentially as for whole animals. Incubation of the enzyme and the antibody was done in a humid chamber; the enzyme reaction was stopped in a water bath. Color reaction was stopped with PBS instead of ethanol and coun-

The Caspase inhibitors z-VAD-FMK and Ac-DEVDCMK (Calbiochem) were dissolved in 100% DMSO and diluted immediately prior to use. Treatments were done with inhibitor concentrations of 1, 10, 100, 300, and 600 ␮M in 0.3– 0.75% DMSO. The solution was replaced every 6 h over the course of 48 h because of the short half-life of these inhibitors (Garcia-Calvo et al., 1998). As a control for unspecific protease effects and unspecific effects by the chloro-methyl-ketone group (CMK) we treated animals with 600 ␮M of N-tosyl-lysine-CMK (Calbiochem), an inhibitor of trypsin-like serine proteases. Peroxidase staining The peroxidase assay was carried out as previously described (Hoffmeister and Schaller, 1985; Technau and Holstein, 1996). Isolation of genomic DNA Fifty polyps or 50 embryos were incubated in lysis buffer (0.2 M Tris, pH 8.0, 100 mM EDTA, 0.2% SDS, 100 ␮g/ml Proteinase K) for 2 h at 60°C; 2 ␮l of diethylpyrocarbonate was added and incubation was continued for another 30 min at 60°C. Next, 50 ␮l of 5 M potassium acetate was added and incubation was carried out on ice for 30 min. The samples were centrifuged for 5 min at 14,000 rpm and the supernatant transferred to a fresh tube and again centrifuged at 14,000 rpm for 15 min. The supernatant was transferred

U. Technau et al. / Developmental Biology 260 (2003) 191–206

into a fresh tube and precipitated by adding 0.75 ml of 100% ethanol and centrifuging for 5 min. The dried DNA pellet was resuspended in 350 ␮l of TE buffer, DNase-free RNaseA was added to 100 ␮g/ml, and the samples were incubated at 37°C for 1 h. The RNase was removed by a phenol– chloroform extraction. The DNA was precipitated with ethanol and resuspended in 50 ␮l of TE buffer. Whole-mount in situ hybridization In situ hybridization was carried out as previously described (Technau and Bode, 1999) except that detection was done with NBT/BCIP instead of BM Purple.

Results Nurse cells phagocytosed by the oocyte are not degraded Ultrastructural studies have shown that nurse cells display several morphological features of apoptotic cells (Zihler, 1973; Honegger et al., 1989). We therefore stained live polyps at different stages of oogenesis with AO. AO binds to nucleic acids (RNA and DNA) in nonfixed apoptotic cells. As the intracellular and nuclear environment become more acidic and chromatin condensation and degradation occur, AO fluorescence shifts from green to a yellow/ red color under blue light excitation (Mpoke and Wolfe, 1997). In Hydra, the first AO-positive nurse cells can be detected at stage 3 of oogenesis, when the oocyte starts phagocytosing the nurse cells (Fig. 1; Miller et al., 2000). During later stages the number of AO-positive nurse cells rapidly increases, and results in a brightly stained oocyte with thousands of densely packed green nurse cells inside the oocyte (Figs. 1A–D). We followed the fate of the nurse cells during embryogenesis and found that the great majority of them can be detected throughout embryogenesis until hatching of the polyp. During this time, their morphology and AO-staining pattern appears virtually unchanged (Figs. 1E– 1). Hence, nurse cells are not degraded after phagocytosis by the oocyte and can be maintained in this nondegraded state. This is remarkable, because hatching can occur up to 12 months after fertilization and probably later (R.E.S., unpublished observations). To further substantiate this observation we macerated a developing oocyte from a female at oogenesis stage 3. When stained with DAPI we found numerous nuclei of the incorporated nurse cells in the oocyte (Figs. 2A and B). The chromatin of these nurse cells is condensed, but the nucleus does not break down into smaller fragments (Fig. 2B). In mature oocytes (stage 7), nurse cell chromatin frequently forms a ring that is displaced to the periphery of the nucleus (Fig. 2C), which is characteristic of cells displaying large-scale fragmentation (see below; Hunot and Flavell, 2001).

195

Caspase inhibitors can block the development of the oocyte Caspases are key factors in the induction of apoptosis in metazoans (Nicholson and Thornberry, 1997). In the canonical apoptotic pathway, activation of caspase 3 ultimately leads to the activation of endonucleases (e.g., DNase 1), which results in the final step in the degradation of apoptotic cells, the internucleosomal fragmentation of nuclear DNA. Two caspases have been isolated from Hydra, termed caspase 3A and caspase 3B. These genes both encode class 3-type caspases, i.e., potential effector caspases (Cikala et al., 1999). We analyzed the mRNA expression of these two caspases during oogenesis. While caspase 3B does not appear to be significantly expressed in the egg field or the developing oocyte (data not shown), caspase 3A expression can be detected at elevated levels in the egg field from stage 2 to stage 4 (Figs. 3A–C). At stage 2, expression is found in the whole egg field (Fig. 3A). At this stage, oocyte selection has just occured (Miller et al., 2000) and the nurse cells have not yet been phagocytosed by the oocyte. We therefore conclude that at an early stage of oogenesis nurse cells express caspase 3A. At oogenesis stage 4, when all nurse cells have been taken up by the oocyte, strongest expression is found in the center of the egg field (Fig. 3B). At stage 5 expression starts to decrease and cannot be detected in oocytes or early embryos (data not shown). This suggests that type 3 caspases might play a role during early oogenesis in Hydra. To test for a functional role of caspases in oogenesis, we treated females at various stages of oogenesis with the broad spectrum caspase inhibitor zVAD-FMK and the caspase 3-specific inhibitor Ac-DEVD-CMK (Garcia-Calvo et al., 1998). When we started the treatment at oogenesis stage 1 or 2 with 0.3– 0.6 mM inhibitor, oogenesis was blocked in a large fraction of females and no mature oocytes were formed (Figs. 4C and D; Table 1). However, when we started the treatment at stage 3 (or later), oogenesis proceeded at the normal rate and the mature oocyte appeared unaffected (Table 1). Also, inhibitor concentrations ranging between 1 and 10 ␮M had no effect; at 100 ␮M they had only minor effects. A control treatment with 0.6 mM Ntosyl-lysine-CMK, an inhibitor of trypsin-like serine proteases, had no effect on the development of oocytes, although it had severe effects on other parts of the polyps (Figs. 4A and B). This suggests that the observed effects of the caspase inhibitors are not due to unspecific effects by the reactive CMK group. Upon early treatment with caspase inhibitors virtually no TUNEL-positive cells could be found in the body column (Figs. 4C and D; Table 1) nor could AO-positive nurse cells be found in the egg field (data not shown). This suggests that caspase activity is necessary for oogenesis during early phases, before phagocytosis of nurse cells, but not required for oocyte development after phagocytosis. Further, the initiation of nurse cell apoptosis appears to be an essential step for proper oocyte development.

196

U. Technau et al. / Developmental Biology 260 (2003) 191–206

Nurse cells phagocytosed by the oocyte are TUNELnegative To test for internucleosomal fragmentation of the genomic DNA in nurse cells, we established a protocol for TUNEL staining in Hydra. As a positive control, we examined hydroxyurea (HU)-treated polyps and polyps undergoing oogenesis. When apoptosis is induced by treatment of polyps with HU, numerous AO- and TUNEL-positive cells can be detected in the body column of the polyp (data not shown). During late stages of oogenesis, a small number of nurse cells located at the periphery of the egg field fail to be phagocytosed by the oocyte. Instead, they are phagocytosed and rapidly degraded by surrounding epithelial cells. When stained with AO, these nurse cells change from green to red fluorescence, indicating advanced DNA degradation (Fig. 5A). Similarly, these nurse cells are TUNEL-positive (Figs. 5B and C). We then performed TUNEL assays during all stages of oogenesis and all accessible stages of embryogenesis (i.e., stages prior to formation of the theca). Surprisingly, nurse cells incorporated by the oocyte were TUNEL-negative at all stages examined, suggesting that fragmentation of the genomic DNA does not occur (see, for example, the postgastrula embryo in Fig. 6A). Thus, although nurse cells phagocytosed by the oocyte are brightly stained green by AO, they never become TUNEL-positive. Nurse cells were also negative for the TUNEL assay when the polyps and early embryos were cut before and after fixation (data not shown), eliminating the possibility that phagocytosed nurse cells are not accessible to the TUNEL assay reagents. Why are nurse cells not degraded inside the oocyte? One possibility is that the nuclear DNA of the nurse cells is somehow protected against cleavage by DNase I. To test this, we treated various stages of oogenesis and embryogenesis with DNase I after fixation, followed by a TUNEL assay. In response to the DNase I treatment numerous epithelial cells of the parent polyp or nuclei of the blastomeres in embryos became TUNEL-positive, yet nurse cells remained TUNEL-negative (Figs. 6B–E). These results support the hypothesis that the DNA of the phagocytosed nurse cells is packed in a form distinct from that of somatic cells, a form which protects it from degradation. DNA of phagocytosed nurse cells is cleaved into large fragments

Fig. 3. In situ hybridization of caspase 3A during oogenesis. (A) stage 2 (B) stage 4 of oogenesis (C) sense control stage 4.

Internucleosomal degradation of nuclear DNA in apoptotic cells is detected by the appearance of fragments in multiples of 200 bp. We analyzed DNA from normal polyps, from polyps induced to undergo apoptosis by HU treatment, from late oogenesis stages, and from early embryos. DNA from normal polyps migrates as a single band at limiting mobility (Fig. 7A). The same band is found in HU-treated animals, presumably derived from unaffected epithelial cells, in addition to bands of 200 and 400 bp,

U. Technau et al. / Developmental Biology 260 (2003) 191–206

197

Fig. 4. Caspase inhibitors can block oocyte formation during the early phase of oogenesis. (A,B) live animals treated with N-tosyl-lysine-chloro-methylketone (TL-CMK), a serine protease inhibitor, which leads to severe damage to the somatic portion of the polyp (A) but has no effect on oocyte formation (B). (C, D) treatment of animals with Ac-DEVD-CMK starting at stage 2 causes arrest of oocyte development. Fig. 5. Peripheral nurse cells that are not phagocytosed by the oocyte become degraded rapidly. (A) AO staining of a stage 7 polyp with a mature oocyte (arrowhead) shows orange-red AO-positive nurse cells at the periphery of the egg field that were not incorporated into the oocyte. (B) A view looking onto the egg cup (the margin of the egg cup is marked by arrowheads) after detachment of the oocyte shows that peripheral nurse cells that were not phagocytosed by the oocyte have become TUNEL-positive (arrows). (C) Close-up of (B) showing TUNEL-positive nurse cells at the periphery of the egg field which have been phagocytosed by epithelial cells. Bars, 200 (A, B) and 60 ␮m (C).

which are likely derived from strongly degraded genomic DNA of the interstitial cells undergoing apoptosis (Fig. 7A). DNA from late oogenesis stages does not differ from nor-

mal polyps and shows no obvious signs of degradation (Fig. 7A). In early embryos, a slight laddering can be detected in some but not all preparations. Intriguingly, the largest sized

198

U. Technau et al. / Developmental Biology 260 (2003) 191–206

Table 1 Effect of caspase inhibitors on oogenesis DMSO control

Ac-DEVD-CMK

z-VAD-FMK

Stage when treatment was started

Stage reached after 48 h of treatment

Tunel⫹ cells

Stage after 48 h

Tunel⫹ cells

Stage reached after 48 h of treatment

Tunel⫹ cells

Stage 2 (n) Stage 3 (n)

5.6 ⫾ 1.4 29 7⫾0 15

220 ⫾ 111 11 n.d. n.d.

2.5 ⫾ 0.6 19 3.5 ⫾ 1.2 11

7.2 ⫾ 3.7 5 n.d. n.d.

1.9 ⫾ 0.7 28 5.3 ⫾ 1.8 16

11 ⫾ 9 9 131 ⫾ 42 5

Note. Females at oogenesis stage 2 and 3 were selected and treated with DMSO (0.75%) or 300 ␮M Ac-DEVD-CMK and z-VAD-FMK. Solutions were changed every 6 h. After 48 h the oogenesis stages were scored and the number of TUNEL-positive cells in the body column was determined.

DNA in the embryo preparations is always noticeably smaller in size than is seen in adult polyps (Fig. 7A, lanes 8 and 9). To analyze this size difference more accurately, we fractionated the DNA on a 0.7% agarose gel. We found that most of the embryo DNA is in fragments of 8 –15 kb (Fig. 7B). The early embryos used for this analysis consisted of 8 –100 blastomeres, but they have incorporated up to 10,000 nurse cells. Hence, these preparations are enriched on the order of 100- to 1000-fold for DNA from the nurse cells. In support of this conclusion, we found that DNA preparations from older embryos with much higher numbers of zygotic cells gave two bands, one at limiting mobility (presumably from the embryo nuclei) and one in the 8- to 15-kb region (presumably from the nurse cell nuclei) (data not shown). These results confirm the negative result of the TUNEL assay. The number of free 3⬘OH DNA ends after the largescale fragmentation is presumably too small to be detected by TUNEL. In conclusion, the DNA in phagocytosed nurse cells does become fragmented, but the fragments are much larger than the internucleosomal fragments seen in cells which are proceeding through the normal course of apoptosis. Persistent peroxidase activity in phagocytosed nurse cells Peroxidases and catalases are known inhibitors of apoptosis, presumably by preventing oxidative stress to mitochondria, a damage which can induce apoptosis (Zhang et al., 1997; Husbeck et al., 2001; Andoh et al., 2002; Petrosillo et al., 2001; Moon et al., 2002; Iwai et al., 2003; Imai et al., 2003; Hockenberry et al., 1993). Peroxidase is therefore potentially an agent responsible for preventing nurse cells from completing apoptosis until hatching. To test this possibility, we performed a peroxidase activity assay at various stages of oogenesis and embryogenesis and on hatchlings. We found a high level of peroxidase activity in the nurse cells from early stages of oogenesis on throughout embryogenesis (Fig. 8A). Newly hatched polyps (Day 0) also exhibit a high level of peroxidase activity (Fig. 8B), which is localized in the phagocytosed nurse cells (Fig. 8C). Interestingly, in many polyps we observe that peroxidase activity in newly hatched polyps decreases within 1–3 days

following hatching, starting at the apical end of the polyp and spreading over the rest of the body (Figs. 8D–F). This pattern of decrease in peroxidase activity is strikingly correlated with the pattern of degradation and precedes the disappearance of nurse cells after hatching (compare Figs. 1G–1), suggesting a role for peroxidase in preventing apoptosis in nurse cells until hatching. Nurse cells are prevented from entering the lysosomal degradation pathway Phagocytosis is an important step in the degradation of apoptotic cells and has been shown to be necessary for completion of apoptosis (Reddien et al., 2001; Hoeppner et al., 2001, Hengartner, 2001). Since phagosomes normally enter the lysosomal pathway by fusion with the acidic secondary lysosomes, we stained oogenesis stages and oocytes with H22.58, a monoclonal antibody specific for late secondary lysosomes (Technau, David, and Holstein, unpublished results). While the antibody readily recognizes lysosomes in the epithelial cells of the same specimen, it does not recognize phagosomes with incorporated nurse cells. H22.58⫹ vacuoles are present in the entire body but absent from the oocyte (Fig. 9A). Similarly, during earlier phases of oogenesis phagocytosed nurse cells are not located in H22.58⫹ lysosomes (Figs. 9B and C). This suggests that nurse cells incorporated by the oocyte do not enter the lysosomal pathway, and thereby escape complete degradation. Although the nurse cells can persist for months throughout embryogenesis in the unhatched polyp, the nurse cells disappear within 3 days after the polyp hatches. The degradation of the nurse cells, which are now all located in the endodermal cell layer in newly hatched polyps, appears to begin in the apical part of the animal and progress down the polyp, ending at the aboral end of the polyp (Figs. 1G–I). Interestingly, this degradation pattern is correlated with a shift of green to orange-red fluorescence of the AO, similar to that seen with i cells that were induced by a treatment with HU to undergo apoptosis (Miller et al., 2000). Since the H22.58 antibody does not penetrate the mesogloea in whole mounts, we prepared macerates from Day 2 and Day 3 hatchlings and stained them with H22.58

U. Technau et al. / Developmental Biology 260 (2003) 191–206

antibody. In endodermal cells of these hatchlings about 10% of the nurse cells are found in the H22.58 secondary lysosomes. DAPI counterstaining shows that all nurse cells in the H22.58 lysosomes are highly degraded, with fragmented nuclei, and can be hardly detected morphologically as nurse cells (Figs. 9D–G). Incorporated nurse cells that can be easily recognized morphologically were never found in the H22.58 lysosomes. This suggests that nurse cells enter the secondary lysosomes only shortly before or during complete degradation and that this step is prevented at earlier stages of embryogenesis. To further confirm this result, we performed a TUNEL assay on macerates of hatchlings and double-stained them with H22.58 antibody. The results clearly showed that in Day 2–3 hatchlings about 99% of all nurse cells were TUNEL-positive, indicating advanced DNA fragmentation (Figs. 9H–J). However, we also still found a small fraction of TUNEL-negative nurse cells and a small number of cells with only peripheral TUNEL activity (Fig. 9J). Interestingly, only TUNEL⫹ nurse cells with highly degraded and fragmented nuclei were found in H22.58 secondary lysosomes (Fig. 9H). These results suggest that fragmentation of the nurse cell DNA is resumed in hatchlings. Finally, the nurse cells enter the lysosomal pathway and are degraded rapidly.

Discussion Apoptosis of nurse cells is arrested upon phagocytosis by the oocyte This study and previous work have shown that Hydra nurse cells display a number of features characteristic of apoptotic cells: the chromatin condenses to one side of the cell; the cells show membrane blebbing, become spherical, and produce membrane-enclosed fragments which are presumably taken up by the growing oocyte (Zihler, 1972; Honegger et al., 1989). Fig. 10 summarizes the fate of the nurse cells during oogenesis and embryogenesis in Hydra. The oocyte becomes determined at oogenesis stage 2, while all other germ cells differentiate into nurse cells (Miller et al., 2000). Shortly thereafter, nurse cells initiate apoptosis, are phagocytosed by the oocyte and become brightly stained by acridine orange following phagocytosis (Figs. 10A and B). During the early stages of embryogenesis, nurse cell DNA becomes fragmented into large pieces about 8 –15 kb in length. Despite all of these changes indicative of apoptosis, the phagocytosed nurse cells are not TUNEL-positive. Further, they are not degraded and do not change much in appearance throughout embryogenesis until hatching, which can take as long as 12 months or longer. In hatchlings, however, they become TUNEL-positive and are eventually rapidly degraded in the secondary lysosomes of the newly hatched polyp. Thus, nurse cells enter the apoptotic pathway, possibly as part of their differentiation program, but completion of

199

nurse cell apoptosis is arrested until hatching of the embryo. Since nurse cells which are not incorporated by the oocyte are phagocytosed and degraded by epithelial cells within one day, we conclude that phagocytosis by the oocyte protects nurse cells from further degradation. This shows that the apoptotic machinery can be stalled by appropriate external signals. Completion of apoptosis, however, resumes quickly after hatching and might involve specific signals from the phagocytes (i.e., the endodermal epithelial cells of the primary polyp). During oogenesis in Drosophila, nurse cells are only completely degraded and resorbed after maturation of the oocyte (McCall and Steller, 1998), although during a much shorter time frame than that seen with Hydra. Another indication that apoptosis can be reversed at intermediate stages comes from studies in C. elegans, which showed that programmed cell death of cells that have already activated ced-3 (caspase 3) can be reverted in mutants deficient for phagocytosis (Hoeppner et al., 2001; Reddien et al., 2001). Blocking phagocytosis in C. elegans leads to higher rates of survival of cells destined to die, suggesting a recovery from initial stages of cell death. In Hydra, phagocytosis by epithelial cells promotes degradation of apoptotic nurse cells. By contrast, phagocytosis by the oocyte leads to an arrest of apoptosis. Honegger et al. (1989) have suggested a cytoplasmic transfer between differentiated nurse cells and the early oocyte before initiation of phagocytosis (stage 3). This nutritional supply by the nurse cells might be necessary for the proper development of the oocyte during early oogenesis. Since degradation of nurse cells is stalled after phagocytosis, what could be the role of nurse cells during late oogenesis and embryogenesis? One possibility is that phagocytosed nurse cells are still alive and continue to synthesize RNA and/or proteins to provide them to the oocyte. However, how the molecules provided to the oocyte are transferred from the endosome into the cytoplasm of the oocyte is completely unclear. Also, it is not clear whether the fragmentation of the nuclear DNA into 8- to 15-kb pieces allows efficient transcription. We therefore favor the idea that nurse cells have only very limited metabolic activity, possibly just enough for survival inside the endosomes. Their main role after phagocytosis would be to act as a source of nutrients for the newly hatched polyp. A caspase-independent pathway of DNA fragmentation in Hydra nurse cells? Two caspase genes have been identified in Hydra and both encode members of the caspase 3 class (Cikala et al., 1999). The level of caspase 3A RNA increases during early oogenesis, but drops in mature oocytes and embryos. Treatment with different caspase inhibitors during early oogenesis can block oocyte formation at high doses but they have no effect if applied during later stages of oogenesis. This suggests an early caspase-dependent phase of nurse cell

200

U. Technau et al. / Developmental Biology 260 (2003) 191–206

Fig. 6. Nurse cells are TUNEL-negative and are not susceptible to DNase I treatment. (A) TUNEL assay on a postgastrula stage embryo shows that nurse cells are TUNEL-negative. (B) TUNEL assay on a postgastrula embryo treated with DNase I after fixation. (C) Close-up view of an early cuticle embryonic stage showing that only embryonic nuclei (arrows) and not the nuclei of incorporated nurse cells (arrowheads) are TUNEL-positive. (D) Animal at oogenesis stage 4 treated with DNase I and showing numerous TUNEL-positive cells. (E) close-up of an egg field of a stage 4 female treated with DNase I shows that only epithelial cells (arrows) but not nurse cells (arrowheads) are TUNEL-positive. Bars, 200 (A, B), 300 (D), and 80 ␮m (C, E).

apoptosis that is necessary for proper oocyte development, and a later caspase-independent phase after phagocytosis, during which apoptosis is arrested. Caspase 3 activation normally leads to the internucleosomal degradation of nuclear DNA by activation of endonucleases such as DNase I. Despite their having many characteristics of apoptotic cells, nurse cells did not show internucleosomal fragmentation even after a long period of incorporation into the oocyte and embryos. Instead, we found degradation into large fragments of about 8 –15 kb. In addition to the formation of large DNA fragments, we also frequently observed the condensation of the chromatin at the periphery of the nuclear envelope which is typical of the situation when DNA degradation produces large fragments (Hunot and Flavell, 2001). Production of such large fragments has been described as an alternative route of apoptosis (Oberhammer et al., 1993; Zhivotovsky et al., 1994; Lagarkova et al., 1995; Susin et al., 2000; Arnoult et al., 2001; Cipriani et al., 2001). This pathway is caspase-independent and involves an unknown nuclease (Hunot and Flavell, 2001). The endonucleases DNase II or DNase ␣ and DNase ␤ have been proposed as possible candidates for this pathway (Shiokawa et al., 1994; Barry and Eastman, 1993; Wu et al., 2000; McIlroy et al., 2000). Unlike DNase I, these enzymes are cation-independent and require an acidic envi-

ronment (Shiokawa et al., 1994; Barry and Eastman, 1993). This fits the observation of a shift from green to orange-red fluorescence in AO-positive nurse cells, which indicates an intracellular acidification (Mpoke and Wolfe, 1997). On the other hand, the recently identified endonuclease G (Endo G) acts as a caspase-independent nuclease, which leads to the late fragmentation of nuclear DNA into pieces of about 50 kb (Parrish et al., 2001; Li et al., 2001; Widlak et al., 2001; Wang et al., 2002). Depletion of Endo G leads to an arrest of apoptosis or to less apoptotic cells, indicating, that Endo G is necessary for the progression of apoptosis. This fits precisely our observation of the DNA cleavage pattern of nurse cells in Hydra. Initiation of the caspaseindependent pathway in vertebrates seems to involve the apoptosis-inducing factor (AIF) that is released from mitochondria and translocates to the nucleus (Joza et al., 2001; Cande et al., 2002). However, AIF also appears to serve as a free radical scavenger, mediating antiapoptotic functions in a manner similar to that of peroxidases (Klein et al., 2002; Lipton and Bossy-Wetzel, 2002). Neither Endo G nor AIF has been reported from Hydra to date, but it will obviously be interesting to investigate the role of these proteins during Hydra oogenesis. The correlation between survival of the nurse cells and

U. Technau et al. / Developmental Biology 260 (2003) 191–206

201

Fig. 7. Analysis of DNA from eggs and embryos in double preparations; (A) 1.5% agarose gel showing DNA from normal polyps (lanes 1, 2), Hydroxyurea-treated polyps (lane 3, 4), oogenesis stages 3–5 (lanes 6, 7) and 8-cell to 100-cell embryos (lanes 8, 9); (B) 0.7% agarose gel of DNA from normal polyps and from early embryos which show degradation into fragments of 8 –15 kb.

strong peroxidase activity suggests that peroxidases might be involved in the arrest of nurse cell apoptosis. Interestingly, peroxidases have been implicated as anti-apoptotic agents in other systems (Zhang et al., 1997; Husbeck et al., 2001; Andoh et al., 2002; Petrosillo et al., 2001; Moon et al., 2002; Imai et al., 2003; Iwai et al., 2003; for reviews see Adrain and Martin, 2001; Salvesen and Renatus, 2002). Induction of apoptosis by oxygen radicals is thought to be an ancient mechanism, already present in yeast (Madeo et al., 1999). Hence, peroxidases might play a crucial role in regulating the arrest of nurse cell apoptosis in the basal metazoan Hydra. Final degradation of nurse cells in secondary lysosomes A striking characteristic of the nurse cell apoptosis is that nurse cells persist in phagosomes without being degraded. However, upon hatching, DNA fragmentation of nurse cells is resumed and the cells enter the secondary lysosomes within 1–3 days. The persistent peroxidase activity provides circumstantial evidence that nurse cells might even have a limited metabolic and synthetic activity. This is reminiscent of a number of intracellular symbionts and pathogens. In Hydra viridissima, the green Hydra, endodermal cells phagocytose symbiotic Chlorella algae. These endosymbionts can proliferate inside the cells and maintain a stable population. The endosymbionts appear to prevent their degradation by secondary lysosomes (Technau, Hegen and Holstein, unpublished results). Likewise, many Anthozoa (corals and sea anem-

ones) harbor intracellular symbiotic Zooxanthella algae. Intracellular parasites and pathogenic bacteria such as Legionella also prevent acidification and fusion of the phagosomes with the lysosomes (Vogel and Isberg, 1999; Amer and Swanson, 2001). Thus, nurse cells in Hydra may use strategies similar to those of endosymbionts and pathogens to survive inside a host cell and may indeed have many features in common with endosymbionts.

Evolution of apoptosis in germline-derived nurse cells Given Hydra’s near basal position in the metazoan tree, it is of obvious interest to try to place our findings in an evolutionary context. Recently, it has been shown that nanos and vasa, two germline-specific genes of Bilateria, are also expressed specifically in germ cells in Hydra (Mochizuki et al., 2000, 2001). This suggests a conserved genetic control mechanism in germline differentiation of all animals (see also Hobmayer et al., 2001). Apoptosis of germline-derived nurse cells also appears to be an ancestral feature of oogenesis (for review see Matova and Cooley, 2001) and phagocytosis of nurse cells by the oocyte has been reported from a number of cnidarian species (Tardent, 1985). Interestingly, some sponges and Ctenophores also show a type of oogenesis very similar to that seen in Hydra (Fell, 1969; Harrison and Westfall, 1991), suggesting that apoptosis and phagocytosis of nurse cells is a basic feature of oogenesis in early metazoans.

202

U. Technau et al. / Developmental Biology 260 (2003) 191–206

Fig. 8. Persistent peroxidase activity in nurse cells throughout oogenesis and embryogenesis. (A) Oogenesis stage 2/3. (B) Newly hatched polyp showing the high level of peroxidase activity. (C) Close-up of the same polyp showing that the peroxidase activity is limited to nurse cells (arrows) which are now all located in the endoderm; (D) 3-day-old hatching. (E, F) Close-ups of the apical and basal regions indicated in (D). Note that most peroxidase-positive nurse cells have disappeared from the apical region, while still a few are found in the basal region of the animal. Bars, 200 (A, B, E, F), 400 (D), and 50 ␮m (C). Fig. 9. Phagocytosed nurse cells are not located in H22.58-positive secondary lysosomes until hatching. (A) Whole-mount staining with mAb H22.58 of female with mature oocyte (arrow). H22.58⫹ vacuoles are present throughout the body column but not in the oocyte. (B, C) View on egg field of oogenesis stage 4 stained with mAb H22.58 (B) and visualized with Nomarski optics (C). Numerous H22.58⫹ lysosomes can be detected in epithelial cells (arrowhead), but the underlying nurse cells (arrow) are not located in such vacuoles. (D–G) macerated epithelial cells of a 3-day-old hatchling stained for H22.58 (green) and DAPI (blue). Note, that only strongly degraded DNA can be found in H22.58 lysosomes (arrows), while more intact nurse cells (arrowheads) are not in lysosomes. (H) Macerated epithelial cells stained for H22.58, DAPI, and TUNEL (black), showing that only strongly degraded TUNEL-positive cells can be found in the secondary lysosomes. Arrows indicate nurse cell remnants located in secondary lysosomes, while arrowheads show less strongly degraded nurse cells which are not yet in the lysosome. The asterisks indicate the nucleus of the epithelial cell. (I–J) TUNEL-stained macerates of Day 3 hatchling, showing that most nurse cells are TUNEL-positive at that stage, but some are still negative (J, left) or appear to become TUNEL-positive at the periphery of the nucleus (I, J, middle). Bars, 200 (A) and 50 ␮m (B, C).

U. Technau et al. / Developmental Biology 260 (2003) 191–206

203

204

U. Technau et al. / Developmental Biology 260 (2003) 191–206

Fig. 10. Summary of nurse cell fate during oogenesis and embryogenesis in Hydra. (A) schematic drawings of the key steps in oogenesis and embryogenesis until formation of the primary polyp, which can be up to 12 months later. The nurse cells are indicated in green, the oocyte and somatic cells of the embryo and polyp in red. Light red corresponds to endoderm, darker red to ectoderm. (B) Progression of apoptosis of nurse cells as indicated by the appearance of several markers or features. See text for further details.

Acknowledgments We thank Charles N. David and Angelika Bo¨ ettger for providing the caspase clones and for fruitful discussions. This work was supported by a grant from the DFG to U.T. (TE 311/3-1) and NSF Grant IBN-9808828 to R.E.S. References Adrain, C., Martin, S.J., 2001. The mitochondrial apoptosome: a killer unleashed by the cytochrome seas. Trends Biochem. Sci. 26, 390 –397. Amer, A.O., Swanson, M.S., 2001. A phagosome of one’s own: a microbial guide to life in the macrophage. Curr. Opin. Microbiol. 5, 56 – 61. Andoh, T., Chock, P.B., Chiueh, C.C., 2002. The roles of thioredoxin in protection against oxidative stress-induced apoptosis in SH-SY5Y cells. J. Biol. Chem. 277, 9655–9660.

Arnoult, D., Tatischeff, I., Estaquier, J., Girard, M., Sureau, F., Tissier, J.P., Grodet, A., Dellinger, M., Traincard, F., Kahn, A., Ameisen, J.C., Petit, P.X., 2001. On the evolutionary conservation of the cell death pathway: mitochondrial release of an apoptosis-inducing factor during Dictyostelium discoideum cell death. Mol. Biol. Cell 10, 3016 –3030. Barry, M.A., Eastman, A., 1993. Identification of deoxyribonuclease II as an endonuclease involved in apoptosis. Arch. Biochem. Biophys. 300, 440 – 450. Bosch, T.C.G., David, C.N., 1984. Growth regulation in Hydra: relationship between epithelial cell cycle length and growth rate. Dev. Biol. 104, 161–171. Cande, C., Cecconi, F., Dessen, P., Kroemer, G., 2002. Apoptosis-inducing factor (AIF): key to the conserved caspase-dependent pathways of cell death? J. Cell Sci. 115, 4727– 4734. Cavaliere, V., Taddei, C., Gargiulo, G., 1998. Apoptosis of nurse cells at the late stages of oogenesis of Drosophila melanogaster. Dev. Genes Evol. 208, 106 –112.

U. Technau et al. / Developmental Biology 260 (2003) 191–206 Cikala, M., Wilm, B., Hobmayer, B., Boettger, A., David, C.N., 1999. Identification of caspases and apoptosis in the simple metazoan. Hydra Curr. Biol. 9, 115–123. Cipriani, B., Borsellino, G., Knowles, H., Tramonti, D., Cavaliere, F., Bernardi, G., Battistini, L., Brosnan, C.F., 2001. Curcumin inhibits activation of Vgammy9Vdelta2 T cells by phosphoantigens and induces apoptosis involving apoptosis-inducing factor and large scale DNA fragmentation. J. Immun. 167, 3454 –3462. David, C.N., 1973. A quantitative method for maceration of Hydra tissue. Wilhelm Roux Arch. Entwicklungsmech. Org. 171, 259 –268. de Cuevas, M., Spradling, A.C., 1998. Morphogenesis of the Drosophila fusome and its implications for oocyte specification. Development 125, 2781–2789. Fell, P.E., 1969. The involvement of nurse cells in oogenesis and embryonic development in the marine sponge Haliclona ecbasis. J. Morphol. 127, 133–150. Foley, K., Cooley, L., 1998. Apoptosis in late stage Drosophila nurse cells does not require genes within the H99 deficiency. Development 125, 1075–1082. Garcia-Calvo, M., Peterson, E.P., Leiting, B., Ruel, R., Nicholson, D.W., Thornberry, N.A., 1998. Inhibition of human caspases by peptide-based and macromolecular inhibitors. J. Biol. Chem. 273, 32608 –32613. Gavrieli, Y, Sherman, Y., Ben-Sasson, S.A., 1992. Identification of programmed cell death in situ via specific labeling of nuclear DNA fragmentation. J. Cell Biol 119, 493–501. Gumienny, T.L., Lambie, E., Hartwieg, E., Horvitz, H.R., Hengartner, M.O., 1999. Genetic control of programmed cell death in the Caenorhabditis elegans hermaphrodite germline. Development 126, 1011– 1022. Harrison, F. W., Westfall, J. A., 1991. Placozoa, Porifera, Cnidaria, Ctenophora, in: Microscopic Anatomy of Invertebrates, Vol. 2, Wiley–Liss, New York. Hengartner, M.O., 2001. Apoptosis: corralling the corpses. Cell 104, 325– 328. Hensey, C., Gautier, J., 1998. Programmed cell death during Xenopus development: a spatio-temporal analysis. Dev. Biol. 203, 36 – 48. Hobmayer, B., Rentzsch, F., Holstein, T.W., 2001. Identification and expression of HySmad1, a member of the R-Smad family of TGFBsignalling transducers, in the diploblastic metazoan. Hydra Dev. Genes Evol. 211, 597– 602. Hockenberry, D.M., Oltval, Z.N., Yin, X.-M., Milliman, C.L., Korsmeyer, S.L., 1993. Bcl-2 functions in an antioxidant pathway to prevent apoptosis. Cell 75, 241–251. Hoeppner, D.J., Hengartner, M.O., Schnabel, R., 2001. Engulfment genes cooperate with ced-3 to promote cell death in Caenorhabditis elegans. Nature 412, 202–206. Hoffmeister, S., Schaller, H.C., 1985. A new biochemical marker for foot-specific cell differentiation in hydra. Roux’s Dev. Biol. 194, 453– 461. Honegger, T.G., Zurrer, D., Tardent, P., 1989. Oogenesis in Hydra carnea: a new model based on light and electron microscopic analyses of oocyte and nurse cell differentiation. Tissue Cell 21, 381–393. Hunot, S., Flavell, R.A., 2001. Death of a monopoly? Science 292, 865– 866. Husbeck, B., Berggren, M.I., Powis, G., 2001. DNA microarray reveals increased expression of thioredoxin peroxidase in thioredoxin-1 transfected cells and its functional consequences. Adv. Exp. Med. Biol. 500, 157–168. Imai, H., Koumura, T., Nakajima, R., Nomura, K., Nakagawa, Y., 2003. Protection from inactivation of the adenine nucleotide translocator during hypoglycema-induced apoptosis by mitochondrial phospholipid hydroperoxide glutathione peroxidase. Biochem. J., 371, 799 – 808. Iwai, K., Kondo, T., Watanabe, M., Yabu, T., Taguchi, Y., Umehara, H., Takahashi, A., Uchiyama, T., Okazaki, T., 2003. Ceramide increases oxidative damage due to inhibition of catalase by caspase-3-dependent proteolysis in HL-60 cell apoptosis. J. Biol. Chem., 278, 9813–9822.

205

Joza, N., et al., 2001. Essential role of the mitochondrial apoptosis-inducing factor in programmed cell death. Nature 410, 549 –554. King, R., 1970. Ovarian Development in Drosophila melanogaster. Academic Press, New York. Klein, J.A., Longo-Guess, C.M., Rossmann, M.P., Seburn, K.L., Hurd, R.E., Frankel, W.N., Bronson, R.T., Ackerman, S.L., 2002. The harlequin mouse mutation down-regulates apoptosis-inducing factor. Nature 419, 367–374. Kleinenberg, N., 1872. Hydra. Eine anatomisch-entwicklungsgeschichtliche Untersuchung, Verlag von Wilhelm Engelmann, Leipzig. Kuznetsov, S., Lyanguzowa, M., Bosch, T.C.G., 2001. Role of epithelial cells and programmed cell death in Hydra spermatogenesis. Zoology 104, 25–31. Lagarkova, M.A., larvaia, O.V., Razin, S.V., 1995. Large-scale fragmentation of mammalian DNA in the corpse of apoptosis proceeds via excision of chromosomal DNA loops and their oligomers. J. Biol. Chem. 270, 20239 –20241. Li, L.Y., Xu, L., Wang, X., 2001. Endonuclease G is an apoptotic DNase when released from mitochondria. Nature 412, 95–99. Lipton, S.A., Bossy-Wetzel, E., 2002. Dueling activities of AIF in cell death versus survival: DNA binding and redox activity. Cell 111, 147–150. Littlefield, C.L., 1991. Cell lineages in Hydra: isolation and characterization of an interstitial stem cell restricted to egg production in Hydra oligactis. Dev. Biol. 143, 378 –388. Madeo, F., Fro¨ hlich, E., Ligr, M., Grey, M., Sigrist, S.J., Wolf, D.H., Fro¨ hlich, K-U., 1999. Oxygen stress: a regulator of apoptosis in yeast. J. Cell Biol. 145, 757–767. Mahajan-Miklos, S., Cooley, L., 1994a. Intercellular cytoplasm transport during Drosophila oogenesis. Dev. Biol. 165, 336 –351. Mahajan-Miklos, S., Cooley, L., 1994b. The villin-like protein encoded by the Drosophila quail gene is required for actin bundle assembly during oogenesis. Cell 78, 291–301. Martin, V.J., Littlefield, C.L., Archer, W.E., Bode, H.R., 1997. Embryogenesis in hydra. Biol. Bull. 192, 345–363. Matova, N., Cooley, L., 2001. Comparative aspects of animal oogenesis. Dev. Biol. 231, 291–320. Matova, N., Mahajan-Miklos, S., Mooseker, M.S., Cooley, L., 1999. Drosophila quail, a villin-related protein, bundles actin filaments in apoptotic nurse cells. Development 126, 5645–5657. McCall, K., Steller, H., 1998. Requirement for DCP-1 caspase during Drosophila oogenesis. Science 279, 230 –234. McIlroy, D., Tanaka, M., Sakahira, H., Fukuyama, H., Suzuki, M., Yamamura, K.-i., Ohsawa, Y., Uchiyama, Y., Nagata, S., 2000. An auxiliary mode of apoptotic DNA fragmentation provided by phagocytes. Genes Dev. 14, 549 –558. Miller, M.A., Technau, U., Smith, K.M., Steele, R.E., 2000. Oocyte development in Hydra involves selection from competent precursor cells. Dev. Biol. 224, 326 –338. Mochizuki, K., Nishimiya-Fujisawa, C., Fujisawa, T., 2001. Universal occurrence of the vasa-related genes among metazoans and their germline expression in Hydra. Dev. Genes Evol. 211, 299 –308. Mochizuki, K., Sano, H., Kobayashi, S., Nishimiya-Fujisawa, C., Fujisawa, T., 2000. Expression and evolutionary conservation of nanos-related genes in Hydra. Dev. Genes Evol. 210, 591– 602. Moon, H., Baek, D., Lee, B., Prasad, D.T., Lee, S.Y., Cho, M.J., Lim, C.O., Choi, M.S., Bahk, J., Kim, M.O., Hong, J.C., Yun, D.J., 2002. Soybean ascorbate peroxidase suppresses Bax-induced apoptosis in yeast by inhibiting oxygen radical generation. Biochem. Biophys. Res. Commun. 290, 457– 462. Mpoke, S., Wolfe, J., 1997. Differential staining of apoptotic nuclei in living cells: application to macromolecular elimination in Tetrahymena. J. Histochem. Cytochem. 45, 675– 683. Nicholson, D.W., Thornberry, N.A., 1997. Caspases: killer proteases. Trends Biochem. Science 22, 299 –306. Nishimiya-Fujisawa, C., Sugiyama, T., 1995. Genetic analysis of developmental mechanisms in hydra. XX. Two types of female germ stem cells

206

U. Technau et al. / Developmental Biology 260 (2003) 191–206

are present in a male strain of Hydra magnipapillata. Dev. Biol. 157, 1–9. Oberhammer, F., Wilson, J.W., Dive, C., Morris, I.D., Hickman, J.A., Wakeling, A.E., Walker, P.R., Sikorska, M., 1993. Apoptotic death in epithelial cells: cleavage of DNA to 300 and/or 50 kb fragments prior to or in the absence of internucleosomal fragmentation. EMBO J. 12, 3679 –3684. Parrish, J., Li, L., Klotz, K., Ledwich, D., Wang, X., Xue, D., 2001. Mitochondrial endonuclease G is important for apoptosis in C. elegans. Nature 412, 90 –94. Petrosillo, G., Ruggiero, F.M., Pistolese, M., Paradies, G., 2001. Reactive oxygen species generated from the mitochondrial electron transport chain induce cytochrome c dissociation from beef-heart submitochondrial particles via cardiolipin peroxidation. Possible role in the apoptosis. FEBS Lett. 509, 435– 438. Reddien, P.W., Cameron, S., Horvitz, H.R., 2001. Phagocytosis promotes programmed cell death in C. elegans. Nature 412, 198 –202. Salvesen, G.S., Renatus, M., 2002. Apoptosome: the seven-spoked death machine. Dev. Cell 2, 256 –257. Seipp, S., Schmich, J., Leitz, T., 2001. Apoptosis—a death-inducing mechanism tightly linked with morphogenesis in Hydractinia echinata (Cnidaria, Hydrozoa). Development 128, 4891– 4898. Shiokawa, D., Ohyama, H., Yamada, T., Takahashi, K., Tanuma, S., 1994. Identification of an endonuclease responsible for apoptosis in rat thymocytes. Eur. J. Biochem. 226, 23–30. Spradling, A.C., 1993. Developmental genetics of oogenesis, in: Bate, M., Martinez-Arias, A. (Eds.), Development of Drosophila melanogaster, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp. 1–70.

Susin, S.A., et al., 2000. Two distinct pathways leading to nuclear apoptosis. J. Exp. Med. 192, 571–579. Tardent, P., 1985. The differentiation of germ cells in Cnidaria, in: Monroy, A. (Ed.), The Origin and Evolution of Sex, A.R. Liss, New York, pp. 163–197. Technau, U., Bode, H.R., 1999. HyBra1, a Brachyury homologue, acts during head formation in Hydra. Development 126, 999 –1010. Technau, U., Holstein, T.W., 1996. Phenotypic maturation of neurons and continuous precursor migration in the formation of the peduncle nerve net in Hydra. Dev. Biol. 177, 599 – 615. Vogel, J.P., Isberg, R.R., 1999. Cell biology of Legionella pneumophila. Curr. Opin. Microbiol. 2, 30 – 4. Wang, X., Yang, C., Shi, Y., Xue, D., 2002. Mechanisms of AIF-mediated apoptotic DNA degradation in Caenorhabditis elegans. Science 298, 1587–1592. Widlak, P., Li, L.Y., Wang, X., Garrard, W.T., 2001. Action of recombinant human apoptotic endonuclease G on naked DNA and chromatin substrates. J. Biol. Chem. 276, 48404 – 48409. Wu, Y.C., Stanfield, G.M., Horvitz, H.R., 2000. NUC-1, a Caenorhabditis elegans DNase II homolog, functions in an intermediate step of DNA degradation during apoptosis. Genes Dev. 14, 536 –548. Zhang, P., Liu, B., Kang, S.W., Seo, M.S., Rhee, S.G., Obeid, L.M., 1997. Thioredoxin peroxidase is a novel inhibitor of apoptosis with a mechanism distinct from that of Bcl-2. J. Biol. Chem. 272, 30615–30618. Zhivotovsky, B., Wade, D., Gahm, A., Orrenius, S., Nicotera, P., 1994. Formation of 50 kbp chromatin fragments in isolated liver nuclei is mediated by protease and endonuclease activation. FEBS Let. 351, 150 –154. Zihler, J., 1972. Zur Gametogenese und Befruchtungsbiologie von Hydra. Wilhelm Roux’s Arch. Entwicklungsmech. Org. 169, 239 –267.