Calcium-regulated fusion of yolk granules is important for yolk degradation during early embryogenesis of Rhodnius prolixusStahl

Chagas disease is the first cause of cardiac lesions in countries of Latin America, where it is endemic (Moncayo,2003), and Rhodnius prolixus is one of the most important vectors of this disease (World Health Organization, 2002). In this context, this insect is crucial for the propagation of Chagas disease and consequently is an important target for vector control. From this point of view, the understanding of reproductive processes, such as oogenesis and embryogenesis, can offer new insights and targets for population control of this insect.

Oogenesis in insects, as in all oviparous animals, occurs by massive incorporations of yolk protein through receptormediated endocytosis(Engelman, 1979; Raikhel and Dadhialla, 1992). Endocytic activities during insect oogenesis involve the incorporation of the lipophosphoglycoprotein vitellogenin(Oliveira et al., 1986; Raikhel and Dadhialla, 1992; Valle, 1993; Sappington and Raikhel, 1998). Once inside the oocyte the vitellogenin, now referred to as vitellin (VT), is stored in organelles known as yolk granules (YGs)(Kunkel and Nordin, 1985; Purcell et al., 1988; Machado et al., 1998). After oviposition, the YGs fill almost the entire volume of the fresh egg and can vary significantly in size and density. During embryogenesis, R. prolixus VT represents the major yolk protein, being mobilized as the main amino acid source for embryo development.

Yolk degradation occurs by activation of acidic hydrolases also stored within YGs. The exact origin of these hydrolases is poorly understood but some of these enzymes are also taken into the oocytes during oogenesis, being stored near yolk proteins in an inactive state. Aedes aegypticarboxypeptidase (Cho et al.,1991) and cathepsin B-like protease(Cho et al., 1999), Blatella germanica vitellin-processing protease and Xenopus laevis lysosomal enzymes have been shown to accumulate in the oocytes during oogenesis (Wall and Meleka,1985; Liu and Nordin,1998; Yin et al.,2001). To activate hydrolases, YGs undergo a process of acidification mediated by proton pumps, such as proton ATPases(H⁺-ATPases) (Fagotto,1991; Nordin et al.,1991; Mallya et al.,1992; Fagotto,1995) and proton pyrophosphatases (H⁺-PPases)(Motta et al., 2004). In insects, several hydrolytic enzymes found in YGs such as cathepsins(Takahashi et al., 1996; Cho et al., 1999; Ribolla et al., 2001), acid phosphatases (Nussenzveig et al.,1992; Ribolla et al.,1993; Fialho et al.,2002; Fialho et al.,2005) and glycosidases(Purcell et al., 1988) were shown to be activated by low pH. Acidification of YGs, therefore, results in activation of hydrolases and degradation of VT. This process is widely conserved within oviparous animals, being essential for yolk mobilization during embryo development. Processing and degradation of VT during embryogenesis of R. prolixus is also dependent on acidification of YGs (Oliveira et al., 1989). In addition, YG-associated hydrolases such as cathepsin D (CD) and acid phosphatase (AP) have been identified in the oocytes of this insect(Nussenzveig et al., 1992; Fialho et al., 2002).

It is generally believed that the YG population is not homogeneous since the vesicles can vary in their macromolecule content and can be fractionated according to their different size and density(Fagotto, 1991; Chestkov et al., 1998; McNeil et al., 2000; Yamahama et al., 2003). For example, in echinoderms, such as starfish and sea urchin, the egg vesicle population is divided in two main groups: reserve and cortical granules. Each group has distinct functions in fertilization signaling and embryo nourishment(Chestkov et al., 1998). In the silkmoth Bombyx mori, the acid enzymes are only present in the small YGs (SYGs), whereas the yolk proteins are localized in the large ones(Yamahama et al., 2003). In the stick insect Carausius morosus, small YGs are frequently more acidic than large YGs and are often seen surrounding the large ones(Fausto et al., 2001). For the hard tick Boophilus microplus, proteolytic activity has also been correlated with differential acidification of YGs and the presence of small vesicles at the periphery of the egg (Abreu et al., 2004). However, the exact origin of the different YGs remains unknown.

Intracellular fusion machinery is composed of dynamic proteins that are assembled upon a signal and are dismantled immediately after the fusion is over, being reutilized. In general, this process is mediated by proteins of the Rab complex (Rabs), SNARES and SM protein families(Jahn et al., 2003). Many proteins involved in membrane fusion are made active by intracellular signaling, involving phosphorylation and protease cleavage(Rutledge and Whiteheart,2002; Lilja et al.,2004; Huyng et al.,2004; Hepp et al.,2005). Calcium binding proteins, such as calmodulin, are known to work as calcium sensors, leading to activation of kinases that in turn activate membrane fusion proteins(Hilfiker et al., 1999). Many membrane fusion events are therefore calcium-dependent and [Ca²⁺]elevation has been described as a fusion signal for several processes(Burgoyne, 1995; Burgoyne and Morgan, 1998). This variation in [Ca²⁺] could be due to Ca²⁺mobilization from specialized internal stores(Tse et al., 1997; Petersen et al., 1999) or influx through plasma membrane (Augustine et al., 1991). In addition, free Ca²⁺ concentration can also be regulated by calcium binding proteins(Meldolesi and Pozzan, 1998). In echinoderms, such as the starfish(McNeil et al., 2000; Chestkov et al., 1998)Ca²⁺ was shown to mediate YG fusion, and Ca²⁺-dependent membrane interactions were recently described in sea urchin eggs(Hayley et al., 2006). In this work, we show that YGs from the blood sucking bug R. prolixus undergo a process of membrane fusion. We show that these events might occur in vivo in a calcium-dependent manner, being important for yolk processing during early embryogenesis of this insect.

Arsenazo III, bovine serum albumin, leupeptin, aprotinin, pepstatin A,dithiothreitol (DTT), phenylmethylsulphonyl - fluoride (PMSF), p-nitrophenyl phosphate (pNPP) and PKH67 Green Fluorescent Cell Linker Kit were purchased from Sigma Chemical Company (St Louis, MO,USA). OCT was obtained from Tissue-Tek (Torrance, CA, USA). All other chemicals were of analytical grade.

Rhodnius prolixus Stahl were reared in a colony maintained at 28°C and 70-80% relative humidity. The animals were fed with rabbit blood in an artificial apparatus according to Garcia et al.(Garcia et al., 1975). Eggs were collected daily and used promptly or allowed to develop to the required embryogenesis stage.

YGs were extracted by gently disrupting the eggs with a plastic pestle in ice cold modified Ringer saline (without addition of CaCl₂)containing 130 mmol l^-1 NaCl, 8.8 mmol l^-1 KCl, 8.6 mmol l^-1 MgCl₂, 10.2 mmol l^-1 NaHCO₃,4.4 mmol l^-1 NaH₂ PO₄, 34 mmol l^-1glucose, pH 7.2 and supplied with a protease inhibitors cocktail (aprotinin,leupeptin, pepstatin A and PMSF).

Approximately 90 eggs, from day 0 or day 3 of development, were disrupted in 500 μl of Ringer saline in the presence of 10 mmol l^-1 EGTA and centrifuged at 50 g for 12 min at 4°C. Pellet and top fractions (enriched in LYGs and SYGs, respectively) were carefully collected,separately resuspended in Ringer saline and centrifuged for 5 min (4°C) at 1000 g, to obtain LYGs, and 17 000 g, for SYGs. The pellets were then resuspended in Ringer saline containing 10 mmol l^-1 EGTA and used in assays.

Opercula from freshly laid eggs or eggs that were allowed to develop until day 3 after oviposition were carefully detached using sharp forceps and a histological blade. The eggs were then fixed in 2.5% glutaraldehyde, 4%paraformaldehyde in PBS for 24 h at 4°C. For cryosections, samples were washed and incubated for 12 h in 20% sucrose in PBS and infiltrated for 96 h in increasing concentrations of OCT (25%, 50%, 75% and pure OCT). After freezing in liquid nitrogen, 7 μm thick transverse sections were obtained,which were adhered to poly-l-lysine-coated glass slides and mounted in glycerol. Alternatively, the fixed material was washed followed by dehydration in an ethanol series (70%, 90% and 2×100%) and embedded in Historesin™ (Leica Historesin Embedding Kit, Nu loch, Heidelberg,Germany). Transverse sections of 7 μm were obtained and stained with 0.1%Toluidine Blue. For LYG counting, 10 different OCT-embedded sections from days 0 and 3 were observed.

Eggs from 0 to 6 days of development were collected and the contents from a pool of four eggs were extracted in 1 ml of modified Ringer saline (pH 7.2). The material was then centrifuged at 10 000 g for 10 min at 4°C and the supernatants incubated in the presence of 4 μmol l^-1 Arsenazo III. Measurements were taken using a CINTRA 20 spectrophotometer (GVC Scientific Equipment Pty Ltd, Dandenog, Australia) at the wavelength pair 675/685. In parallel, calcium titers were added to modified Ringer saline and measured in the presence of 4 μmol l^-1 Arsenazo III (Yingst and Hoffman, 1983).

Day 0 eggs were extracted in 50 μl of modified Ringer saline containing 10 mmol l^-1 EGTA (for chelating of endogenous Ca²⁺) and incubated in increasing concentrations of CaCl₂ (15 mmol l^-1, 17 mmol l^-1 and 27 mmol l^-1), at 25°C. Because endogenous [Ca²⁺] can vary among individuals, the Ca²⁺ concentration for each experimental group was measured and adjusted to achieve final calcium concentrations of 6.6±1.8 mmol l^-1, 11.6±3.1 mmol l^-1 and 23.4±2.2 mmol l^-1 (mean ± s.d.?). After 1 min, 5 μl of each experimental group were deposited on glass slides and observed using differential interference contrast (DIC) light microscopy. The YGs in one random field of each slide were measured and counted. We considered large YGs(LYGs) as structures >40 μm in diameter and small YGs (SYGs) as structures <10 μm in diameter. Statistical analyses were performed using Graph Pad Instat 4.0 software (P=0.05). Results are representative of at least three experiments.

SYGs membranes were labeled with PKH67 Green Fluorescent Cell Linker Kit(Sigma Chemical Co.) according to the manufacturer's protocol. Aliquots of labeled fractions were observed in a fluorescence microscope to assure their purity. For fusion assays, 5 μl of the labeled SYGs were mixed with 20μl of day 0 total YG fraction and incubated in the presence of 23 mmol l^-1 CaCl₂. The material was observed in a Zeiss Axioplan epifluorescence microscope equipped with a fluorescein filter set and a TK-1270 JVC color video camera.

LYGs and SYGs were obtained and their protein concentration determined by the Lowry method (Lowry et al.,1951). 15 μg of protein of each sample were subjected to SDS-PAGE 7.5% (Laemmli, 1970). The gel was stained with silver nitrate(Merril et al., 1981).

Day 0 and 3 eggs were fixed in 4% paraformaldehyde in PBS, pH 7.2, for 24 h at 4°C followed by infiltration in OCT and sectioning as described above. Sections were blocked for 30 min in 100 mmol l^-1 NH₄Cl in TBS, washed and incubated for 30 min in 3% BSA in TBS plus 0.8% Triton X-100 at room temperature. Sections were then incubated for 1 h in polyclonal antibodies raised against R. prolixus vitellins diluted 1:100 in TBS 3% BSA, washed and incubated for 1 h at room temperature in Cy3-conjugated anti-rabbit IgG secondary antibodies. Samples were then washed again, mounted in N-propyl gallate and observed in a Zeiss Axioplan epifluorescence microscope equipped with a rhodamine filter set and a TK-1270 JVC color video camera.

For the determination of acid phosphatase (AP) specific activity, SYGs and LYGs were subjected to three cycles of freeze and thaw and centrifuged at 20 000 g for 30 min. Protein concentrations of the supernatants were determined by the Lowry method (Lowry et al., 1951). For each fraction, aliquots containing 30 μg of protein were assayed at 37°C against 4 mmol l^-1pNPP in the following reaction medium: 20 mmol l^-1 sodium acetate, pH 4.0, 1 mmol l^-1 DTT and 1 mmol l^-1 EDTA. Reactions were stopped after 1 h by the addition of 0.2 mol l^-1 NaOH(corresponding to 10% of the total reaction volume) and each sample,containing the reaction hydrolysis product (p-nitrophenol, pNP), had their absorbance measured at 405 nm in a Thermomax microplate reader (Molecular Devices, Sunnyvale, CA, USA). For cathepsin D(CD) specific activity, SYGs and LYGs were submitted to freeze and thaw,centrifuged as described above and assayed using the CD-specific fluorogenic peptide substrate Abz-AIAFFSRQ-EDDnp(Pimenta et al., 2001). For each fraction, 300 μg of protein were incubated with 5 μmol l^-1 fluorogenic peptide diluted in 20 mmol l^-1 sodium acetate, pH 4.0, 1 mmol l^-1 DTT and 1 mmol l^-1 EDTA at 37°C for 10 min. The fluorescent products were monitored with an F-max fluorometer (Molecular Devices) using a 320 nm excitation filter and 460 nm emission filter.

Membrane fractions of SYGs and LYGs were obtained as follows: samples were resuspended in equal volumes of icecold buffer containing 10% (v/v) glycerol,0.13% (w/v) BSA, 5 mmol l^-1 EDTA, 150 mmol l^-1 KCl, 3.3 mmol l^-1 DTT, 1 mmol l^-1 PMSF and 100 mmol l^-1 Tris-HCl, pH 8.0. The YG suspensions were homogenized using a glass Potter-Elvehjem homogenizer and then centrifuged at 10 000 g for 20 min at 4°C. The supernatants obtained were centrifuged at 100 000 g for 40 min at 4°C and the pellet was resuspended in 10% glycerol, 10 mmol l^-1 Tris-HCl, 1 mmol l^-1 EDTA and 1 mmol l^-1 DTT, pH 7.5 and recentrifuged at 100 000 g for 40 min at 4°C. The final pellet was resuspended in a small volume of the last buffer and assayed for H⁺-PPase and vacuolar H⁺-ATPase activities(Motta et al., 2004). Protein concentration was determined as described above. Reactions were started by the addition of 40 μg of protein in reaction medium containing 50 mmol l^-1 Mops-Tris, 100 mmol l^-1 KCl, 0.3 mmol l^-1NaPPi (sodium pyrophosphate), 0.6 mmol l^-1 MgCl₂, pH 7.5 for H⁺-PPase and 50 mmol l^-1 Mops-Tris, 100 mmol l^-1 KCl, 1 mmol l^-1 ATP, 2 mmol l^-1MgCl₂, pH 7.5, for vacuolar H⁺-ATPase. After 1 h at 28°C, reactions were stopped by the addition of 50 μl of trichloroacetic acid (50% w/v) and colorimetrically measured by determining the rate of Pi release (Fiske and Subbarow, 1925). For vacuolar H⁺-ATPase, all samples were also incubated in the presence of 50 mmol l^-1 nitrate(KNO₃), which is a specific vacuolar H⁺-ATPase inhibitor(Moriyama and Nelson, 1989). Only the nitrate inhibited activity was considered. Statistical analyses were performed in Graph Pad Instat 4.0 software (P=0.05). Results are representative of at least three experiments.

To investigate the location of proton pumps (H⁺-PPase and vacuolar H⁺-ATPase), day-0 and day-3 eggs were extracted and incubated in the dark for 10 min in Ringer saline plus 10 mmol l^-1EGTA containing 5 μg ml^-1 Acridine Orange (AO). For experimental groups, PPi and ATP were added for final concentrations of 0.3 mmol l^-1 and 1 mmol l^-1, respectively. After incubation the YGs were deposited on glass slides and observed at an excitation wavelength of 418 nm in a Zeiss Axioplan epifluorescence microscope equipped with a fluorescein filter set and a TK-1270 JVC color video camera.

YGs were extracted from day-0 and day-3 eggs and incubated in Ringer saline plus 10 mmol l^-1 EGTA containing 1 mmol l^-1 ATP and 1 mmol l^-1 PPi. For experimental groups ∼23 mmol l^-1Ca²⁺ was added. The samples were incubated for 1 h at 37°C and 13 μg of protein, for each sample, were submitted to 10% SDS-PAGE. The gel was stained with silver nitrate (Merril et al., 1981).

Transverse sections revealed differences in the internal morphology between eggs from day 0 and day 3 of development. In day-0 eggs, the YG population filled the entire egg, whereas in day-3 eggs the YG population was variable in size and LYGs were often more abundant(Fig. 1A,B, white arrows). A visible embryo, at the gastrula stage, could be observed only during day 3 of development (Fig. 1B). To differentiate between YGs and lipid droplets, the material was also prepared for Historesin™, a procedure in which the material is submitted to a dehydration series in organic solvents and part of the lipid component is extracted leaving the protein content intact. Fig. 1C,D shows that day-3 eggs had a higher number of LYGs (black arrows), as observed in DIC images of unstained material. Morphometric analysis showed that LYGs are sixfold more frequent in eggs during day 3 than during day 0 of early embryogenesis. The number of LYGs per egg section were: 2±1 on day 0 and 12±2 on day 3 (mean ± s.d.).

Calcium concentration was determined in eggs from 0 to 6 days of development. Fig. 2 illustrates an increase in free calcium concentration in eggs from 0 to 3 days of development. [Ca²⁺] remained at ∼7 mmol l^-1 during days 0 and 1, beginning to increase on day 2 (∼12 mmol l^-1). The highest free [Ca²⁺] was found to be 23±2.3 mmol l^-1 on the third day of embryogenesis. This period is coincident with the beginning of the yolk mobilization in Rhodnius prolixusembryogenesis (Fialho et al.,2005). After that, [Ca²⁺] continuously decreased until day 5 and remained at similar levels (∼18 mmol l^-1) until the sixth day of embryogenesis.

Fig. 3A shows that incubation of YGs obtained from day 0 eggs with ∼7 mmol l^-1Ca²⁺ (the amount of Ca²⁺ found in day-0 and -1 eggs) led to the appearance of threefold more LYGs in comparison with those incubated with EGTA. Incubation of day 0 YGs with the [Ca²⁺] found in day-2 eggs (∼12 mmol l^-1) increased the number of LYGs sevenfold and,∼23 mmol l^-1 Ca²⁺, found in day-3 eggs, induced the formation of 13-fold more LYGs in day 0 eggs. LYG formation showed a dose-response profile to Ca²⁺ addition. Treatment of day 0 YGs with[Ca²⁺] over 23 mmol l^-1 did not result in any significant LYG increase (data not shown). In Fig. 3B a general view of day-0 YGs in the presence of 10 mmol l^-1 EGTA (calcium-free conditions)is illustrated, and in Fig. 3Cday-0 YGs in the presence of ∼23 mmol l^-1 Ca²⁺ is shown. In the latter, the appearance of LYGs is clearly evident. Because day-3 eggs contained the highest [Ca²⁺] found during embryogenesis and this concentration (∼23 mmol l^-1) is the most effective in inducing LYG formation, all subsequent experiments were performed in the presence of ∼23 mmol l^-1 Ca²⁺, and day-0 and day-3 eggs were compared.

After centrifugation, two fractions of YGs, enriched in either LYGs or SYGs, were obtained (Fig. 4)and they were then labeled with the lipophilic fluorescent membrane marker PKH67 and incubated with calcium. Incubation of SYGs with PKH67 resulted in labeling of the membranes (Fig. 5A,B). To investigate if the increase in the amount of LYGs found in day-0 eggs incubated in the presence of calcium was the result of calcium-induced membrane fusion, PKH67-labeled SYGs was added to day 0 total YG fraction (without labeling) in the presence of 23 mmol l^-1Ca²⁺ (the calcium concentration found in day-3 eggs). Fig. 5C,D shows that incubation with calcium induces the transfer of the dye from SYGs to LYGs. When the labeled SYGs were incubated with day 0 total YG fraction in the presence of 10 mmol l^-1 EGTA (no free calcium), no transfer of the dye from SYGs to LYGs could be observed (Fig 5E,F).

Membrane fractions of SYGs and LYGs were separated as described in the Materials and methods and assayed for H⁺-PPase and vacuolar H⁺-ATPase-specific activities. The results showed that SYGs from days 0 and 3 contain similar levels of H⁺-PPase activity but higher levels than LYGs (Fig. 6A). By contrast, LYG H⁺-PPase activities showed differences between day 0 and day 3 with a 90% increase in H⁺-PPase activity on day 3(Fig. 6A). Because egg homogenate activity levels did not change, i.e. it seems that there is no H⁺-PPase synthesis between days 0 and 3, this suggests that the increase in LYG H⁺-PPase activity could be the result of transfer of the enzyme from SYGs to LYGs. To investigate whether this increase in H⁺-PPase activity was the result of calcium-mediated events, day-0 eggs were incubated in the presence of 23 mmol l^-1 Ca²⁺and the H⁺-PPase activity was measured. Results showed that incubation with calcium increased the H⁺-PPase activity in day-0 LYGs (Fig. 6B), suggesting that the calcium-induced events observed in vitro (Figs 3 and 5) are also able to cause increase in H⁺-PPase activity in LYGs. To further investigate whether the H⁺-PPase could be relocated during embryogenesis, YGs were incubated with PPi in the presence of AO, which has been proved to be a reliable marker of intracellular acidic compartments(Anderson and Orci, 1988), and observed with a fluorescence microscope. Results showed that addition of PPi preferentially induced the acidification of SYGs on day 0(Fig. 6C,D) whereas on day 3 it also induced the acidification of large ones(Fig. 6E,F). During embryogenesis the YGs undergo endogenous acidification, thus incubation of day 3 YGs with AO in the absence of PPi showed SYGs and LYGs already acidified(not shown). To confirm if PPi addition was able to induce further acidification of LYGs, acidified LYGs from day-3 eggs incubated, or not, with PPi were counted and compared. Results showed that PPi addition increased by 18% the number of acidic LYGs during day 3 of embryogenesis, indicating that PPi addition also induces LYG acidification during day 3.

In contrast to the previous finding with the vacuolar H⁺-PPase,the vacuolar H⁺-ATPase activity did not significantly alter between days 0 and 3 of development (Fig. 7A). Incubation of YGs with ATP in the presence of AO induced acidification of both SYGs and LYGs at both stages of development(Fig. 7B-E).

SYGs and LYGs were assayed for determination of acid phosphatase (AP) and cathepsin D (CD)-specific activities. It has been shown that day-0 egg homogenates contain very low levels of CD(Fialho et al., 2005) and AP(Fialho et al., 2002)activities. Thus, to better understand the data, both samples were normalized and expressed as a percentage of the total activity of the egg homogenate(relative activity) for each day. Results showed that day-0 SYGs contained the highest levels of AP activity, twofold higher than LYGs(Fig. 8A). However, on day 3 this profile was strongly altered and LYGs showed 30% more activity than SYGs. As observed for H⁺-PPase, this data suggests that LYG AP activity increases could possibly be achieved by enzyme relocation mediated by transfer from SYGs to LYGs. By contrast, as observed for vacuolar H⁺-ATPase,no significant differences in the levels of CD activity between day 0 and 3 of embryogenesis or between SYGs and LYGs were observed(Fig. 8B).

To analyze the protein profile of different YG fractions, LYGs and SYGs were submitted to SDS-PAGE. Fig. 9A reveals that vitellin apoproteins are equally present in egg homogenate, SYGs and LYGs. Arrows indicate the four apovitellins, as described elsewhere (Masuda and Oliveira,1985). Immunofluorescence analysis using antibodies against these proteins showed the presence of vitellins in both LYGs and SYGs on days 0 (Fig. 9B,C) and 3(Fig. 9D,E) of embryogenesis,confirming the previous results with SDS-PAGE. To investigate whether incubation with calcium could have an effect on the VT apoprotein degradation in the eggs from different days, day-0 and -3 YGs were submitted to degradation assay in the presence, or absence, of 23 mmol l^-1calcium. EGTA-treated day-0 YGs showed no degradation whereas incubation with 23 mmol l^-1 Ca²⁺ led to mild VT proteolysis(Fig. 10A, lanes 1 and 2,respectively). Because day 3 of embryogenesis is the period when R. prolixus VT proteolysis starts(Fialho et al., 2005),incubation of day-3 YGs in the presence of EGTA presented a typical endogenous VT proteolysis profile (lane 3), which shows more proteolysis-derived fragments than day-0 eggs incubated in the presence of EGTA or calcium (lanes 1 and 2). Even so, incubation of day-3 YGs in the presence of calcium still resulted in an increase of VT proteolysis (lane 4). Because both ATP and PPi were added to all samples (EGTA and calcium), this data suggests that acidification of YGs per se is not sufficient to activate all yolk degradation machinery, indicating that the observed calcium-induced events take part in this process, probably allowing the assembly of the yolk degradation machinery. To exclude the possibility that VT degradation was the result of calcium activation of CD activity, EGTA and Ca²⁺ samples were assayed for CD activities as described before(Fig. 10B). Results showed that calcium addition did not have any significant effect on increasing CD activity, which suggests that the induction of VT proteolysis was the result of the previously observed calcium-induced events, probably the YG fusions.

It is generally accepted that, in insects, yolk degradation occurs at the onset of embryonic development. In this process VT undergoes proteolysis,triggered by activation of latent proproteases(Fagotto, 1990; Nordin et al., 1991). Activation of these enzymes is achieved by luminal acidification of YGs, a process that is mediated by proton pumps(Fagotto, 1991; Nordin et al., 1991; Mallya et al., 1992; Fagotto, 1995). VT and hydrolases, such as cathepsin B and carboxypeptidase, are known to accumulate in the oocytes during oogenesis and to be stored in YGs(Wall and Meleka, 1985; Oliveira et al., 1986; Raikhel and Dadhialla, 1992; Valle, 1993; Sappington and Raikhel, 1998; Cho et al., 1991; Cho et al., 1999; Liu and Nordin, 1998; Yin et al., 2001). Newly laid eggs are therefore able to initiate VT degradation. In vitro,measurements of hydrolase activities and yolk degradation using fresh eggs have been investigated by many groups(Nussenzveig et al., 1992; Logullo et al., 1998; Sorgine et al., 2000; Fialho et al., 2002). However,even though freshly laid eggs carry all the yolk mobilization machinery, the massive yolk proteolysis does not occur in vivo immediately after the egg is laid, but starts at a certain time later, during early embryogenesis. In R. prolixus, massive yolk proteolysis starts at day 3 of embryogenesis (Fialho et al.,2005).

Fusion of YGs during oviparous embryogenesis has been suggested and described by several groups (Nordin et al., 1991; Chestkov et al.,1998; McNeil et al.,2000; Yamahama et al.,2003) and the impressive capacity of YGs to form large structures by homotypic fusion is well illustrated by formation of large membrane barriers in sea urchin eggs (McNeil et al., 2000). Calcium has been shown to mediate YG fusion, in vitro, in the starfish Asterina miniata, the sea urchin Lytechinus pictus and the insect Periplaneta americana(Chestkov et al., 1998; McNeil et al., 2000; Ramos et al., 2006). Numerous egg responses, such as production of the fertilization envelope, utilize Ca²⁺ as a messenger (Steinhardt et al., 1977).

In this work, we suggest that YGs undergo membrane fusion in vivo,in a calcium-dependent manner, during the third day of embryogenesis in the blood sucking insect Rhodnius prolixus. This process is followed by mobilization of vacuolar H⁺-PPase, acid phosphatase activation and VT proteolysis. Because of the large amounts of calcium ingested in the R. prolixus diet, it is known that several metabolic processes operate with high calcium concentrations in this insect. Rhodnius hemolymph contains 1-8 mmol l^-1 of calcium, whereas the Malpighian tubules can store calcium at concentrations close to 1 mol l^-1(Maddrell et al., 1991). The origin of calcium inside the eggs is unknown, but a calcium-binding protein has been described for R. prolixus and shown to play some role on its transport to the oocytes during oogenesis(Silva-Neto et al., 1996). The finding of such high amounts of calcium inside the egg could be explained by the existence of calcium stores. Since the egg is filled with vesicles, it would be relatively easy for this type of system to operate with calcium pumps, channels and/or exchangers, modulating the amount of free calcium remaining in the ooplasma or inside compartments. Variation of[Ca²⁺] during R. prolixus embryogenesis could also be explained by the possible presence of high amounts of calcium-binding proteins, such as calmodulin, modulating its availability. The presence of calmodulin has been reported in B. germanica eggs, corresponding to 1.5% of the total volume of its soluble proteins, which can be reduced to almost undetectable levels at the beginning of yolk protein cleavage(Zhang and Kunkel, 1992),suggesting a connection between Ca²⁺ availability and yolk degradation. This is in agreement with our observations that the highest[Ca²⁺] is found on the third day of development, the time when yolk degradation starts in R. prolixus eggs(Fialho et al., 2005). Moreover, elevation of [Ca²⁺] represents a controlled way to regulate the YG fusion events. Calcium signaling during embryonic development has been elegantly reviewed (Webb and Miller, 2003), and shown to take part in fertilization, embryonic cleavage, blastula formation and other embryogenesis events. Yolk degradation during embryogenesis is probably modulated by the embryo demands. In this regard, Ca²⁺ signaling inside the egg could also be regulated by the growing embryo.

Whether or not [Ca²⁺] varies between different egg locations (by local release from internal stores and/or association with calcium binding proteins) is still uncertain. Unfortunately R. prolixus eggs contain an opaque and impermeable eggshell, precluding efficient clarification or permeability increase. Thus, experiments involving direct calcium imaging in the ooplasm in vivo are still not possible in this model. However,even though we cannot access the organization of global or local calcium signaling in eggs of R. prolixus, the presence of calcium wave pacemakers and local calcium signaling (as calcium puffs) in eggs(Dumollard et al., 2002) and oocytes (Parker and Yao, 1995)from other models support our findings that Ca²⁺ elevation during early embryogenesis can trigger the fusion of YGs as a slow calcium release on the ooplasm.

To start yolk degradation, the YGs must contain at least three components:yolk proteins, proton pumps and hydrolases. If one of these three components is not found in a common vesicle, yolk degradation does not proceed. As highlighted in this paper, the YG population is not homogeneous and it has been extensively shown that YGs can vary in size, density, macromolecule content, and that not all of them undergo acidification during embryogenesis(Postlethwait and Giorgi,1985; Wallace,1985; Fagotto,1995; Chestkov et al.,1998; McNeil et al.,2000; Fausto et al.,2001).

Our findings that [Ca²⁺] elevation in vivo is coincident with the appearance of more LYGs, combined with the fact that LYG formation and membrane label transfer could be achieved in vitroafter Ca²⁺ treatment provide evidence for the suggestion that R. prolixus YGs undergo a process of membrane fusion. LYG formation observed at day 3 could also be the result of SYGs fusing together, and participation of medium sized YGs in this process cannot be disregarded. However, scanning electron microscopy observations of day 3 YGs commonly evidenced SYGs in association with LYGs (data not shown), suggesting that this type of fusion event might occur more frequently. The observation of higher H⁺-PPase activity in SYGs at the beginning of embryogenesis and the elevation of this activity in LYGs during the third day of development, when the massive YG fusion events are taking place, in combination with an increase of AP activity, suggest the presence of a functional mechanism of macromolecule transference from SYGs to LYGs during the third day of development. This is consistent with this being the time when yolk mobilization begins in R. prolixus embryogenesis(Fialho et al., 2005).

However, it is important to emphasize that H⁺-ATPase and CD levels were unaltered between days 0 and 3, being very similar within the different YG fractions, indicating that these enzymes are not affected by calcium-induced events. Fialho et al.(Fialho et al., 2005) showed that Rhodnius CD and AP are involved in VT degradation in a process in which CD is only able to induce yolk degradation after previous VT dephosphorylation, mediated by AP. This model of modulation of yolk degradation is consistent with our findings that only AP seemed to be transferred by fusion events and could be used for modulation of yolk degradation. Analysis on VT location showed that it seems to be present in all YGs and this pattern does not alter between days 0 and 3, which suggests that VT location is independent of calcium-induced events.

A basic feature of eukaryotic cells is that compartmental organization works through regulated mechanisms of macromolecule sorting to their appropriate compartment. In this context, membrane fusion is a fundamental cellular process, being indispensable for compartmental organization(Alberts et al., 2001).

Altogether, the results indicate that calcium-induced events could have a role in the transfer of components of yolk machinery, regulating the access of hydrolases and/or proton pumps to the yolk proteins and modulating their degradation. This hypothesis is further supported by the finding that VT proteolysis is stimulated by calcium-induced events. Incubation of day 0 YGs in the absence of Ca²⁺ resulted in no VT proteolysis, even when acidified in the presence of ATP and PPi. During day 3, where the membrane fusion events take place, VT proteolysis (higher than on day 0) could also be stimulated by calcium. As Ca²⁺ did not induce any significant increase in CD activity, VT degradation seems to be achieved after YG fusion events, suggesting that it brings all components together, allowing the assembly of the yolk degradation machinery.

Our findings, therefore, provide evidence that calcium-regulated YG membrane fusion is a potential mechanism for macromolecule transfer between different compartments inside the egg. This process probably works on modulation of yolk degradation during embryogenesis, which is essential for VT degradation. Understanding exactly how this process occurs could clarify the dynamics of yolk degradation during insect embryo development. Further studies on this mechanism, such as the presence of calcium binding proteins (sensors and buffers) and calcium stores inside the egg, isolation and location of enzymes involved with yolk degradation and characterization of proteins involved with membrane fusion are currently under intense investigation by our group.

List of abbreviations
 
• AP

  acid phosphatase

 
• CD

  cathepsin D

 
• DIC

  differential interference contrast

 
• LYG

  large YG

 
• NaPPi

  sodium pyrophosphate

 
• SYG

  small YG

 
• VT

  vitellin

 
• YG

  yolk granule

We express our gratitude to Hatisaburo Masuda, Hector Barrabin and Ulysses Lins for kindly providing laboratory supplies and facilities, to Litiane Silva and José de S. L Junior for excellent technical assistance, and to Marcelo Medeiros for critical reading of this manuscript. This work was supported by grants from the following Brazilian agencies: Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq),Fundação Carlos Chagas Filho de Amparo a Pesquisa (FAPERJ),Programa de Nanociencia e Nanotecnologia MCTCNPq and Programa de Apoio ao Desenvolvimento Científico e Tecnológico (PADCT),FINEP/PRONEX/FUJB (no. 76.97.1000.000).