Arginine kinase in the demosponge Suberites domuncula:regulation of its expression and catalytic activity by silicic acid

Sponges (phylum Porifera) are sessile filter feeders that in some marine biotopes represent the most dominant metazoan taxon on the benthos. Their growth rate is remarkably high, but only a few exact measurements have been published. For some species, e.g. Haliclona loosanoffi, an increase in size during the summer period of up to tenfold has been described(Hartman, 1958). It is therefore conceivable that they consume considerable amounts of organic matter to maintain their metabolism. Some sponges filtrate 0.002-0.84 cm³of water s^-1 cm^-3 of sponge tissue through their aquiferous canal system (see Osinga et al., 2003). Calculated from this immense filtration capacity,sponges such as Pseudosuberites andrewsi take up approx. 3 million particles (mostly algae) per cm³ tissue min^-1 from the surrounding aqueous milieu (Osinga et al.,2001). The particle/food uptake and consumption depends strongly on the ambient physical conditions, i.e. air, light, oxygen, salinity and temperature (Osinga et al.,1999; Gatti et al.,2002). But another factor, silica, also strongly influences the growth rate, the form of the spicules and the body size of these animals. Sponges from the classes Demospongiae and Hexactinellida possess a skeleton composed of polymerized silicic acid, silica, while the third sponge class,Calcarea, build their skeleton from calcium carbonate, a salt that is, in contrast to silicic acid, usually not rate limiting in the aqueous milieu.

Silica contributes over 60% of the dry demosponge biomass(Desqueyroux-Faundez, 1990),sometimes even reaching 90% (Barthel,1995). Since the average concentration of silica in seawater is low, with values of less than 3 μmol l^-1(Maldonado et al., 1999) close to the surface, it can be extrapolated that most of the energy generated by sponges is required to build the siliceous skeleton. This assumption is supported by a recent study, which demonstrated that sponges are provided with a (potential) silicic acid transporter that depends on the establishment of energy-requiring ion gradients; this (potential) transporter comprises a 4,4-diisothiocyanatostilbene-2,2-disulfonic acid (DIDS) binding site(Schröder et al., 2004a). DIDS prevents uptake of silicic acid in sponge cells, indicating that the(potential) silicic acid transporter is involved in the formation of silica in the spicules. The silica of the spicules is synthesized enzymatically with the help of silicatein (Cha et al.,1999; Krasko et al.,2000). The substrate required for silicatein to form silica is actively taken up by the tissue and finally by the cells, sclerocytes, which form the spicules (Schröder et al.,2004a).

Since the studies of Roche et al.(Roche and Robin, 1954; Roche et al., 1957) it has been well established that sponges contain the creatine phosphate/creatine kinase system. These findings were confirmed by molecular biological studies on the demosponge Tethya aurantia(Ellington, 2000; Sona et al., 2004). Based on their data it was hypothesized that the creatine kinase system evolved early in metazoan evolution in organisms composed of highly polarized cells(Ellington, 2001). In line with these studies, we extend the available biochemical/molecular biological data for sponges to formulate the pathway involved in the formation of ATP in the cytoplasm.

In eukaryotic cells the major portion of ATP is generated in the mitochondria via oxidative phosphorylation; the high energy equivalents have then to be transported to the cytoplasm. The enzymes that catalyze the reversible transfer of high-energy phosphoryl groups of ATP to naturally occurring guanidine compounds, e.g. creatine, glycocyamine,taurocyamine, lombricine and arginine, are the phosphagen kinases (see Suzuki et al., 1997). In vertebrates, phosphocreatine is the only phosphagen, and the corresponding phosphagen kinase is creatine kinase, while in invertebrates more kinases corresponding to the respective phosphagen have been described (see Muhlebach et al., 1994). On the grounds of the high sequence similarity among these kinases it has been proposed that members of the phosphagen kinases evolved from one common ancestor (Suzuki and Furukohri,1993); however, to date the evolutionary processes are not fully understood (Ellington, 2001). Phosphagen kinases are missing in fungi and plants.

In the present study we show that the gene encoding arginine kinase is differentially highly expressed during exposure of sponges to silicic acid. We identified the cDNA from the demosponge Suberites domuncula and determined its expression level. In parallel, the enzymic activity of arginine kinase was monitored to establish the importance of this enzyme during siliceous spicule formation. S. domuncula is suitable for such experimental studies since it can be maintained in aquaria for over 3 years(Le Pennec et al., 2003) and proliferating cells from this sponge can be cultivated in 3D-cell cultures,termed primmorphs (Müller et al.,1999); furthermore an Expressed Sequence Tags (EST) library containing over 15 000 cDNAs is available(http://spongebase.genoserv.de/). Searching this database revealed no creatine kinase or other phosphagen kinases besides the arginine kinase.

The results reported here suggest that silicic acid causes an induction of the expression of the arginine kinase gene and an increase of enzyme activity in primmorphs after incubation. This effect could be abolished by coincubation with DIDS. Recently it was found in sponge cells that DIDS blocks the uptake of silicic acid (Schröder et al., 2004a).

The sources of chemicals and enzymes used were as given previously(Kruse et al., 1997; Krasko et al., 2000). Natural sterile filtered seawater, hexokinase, glucose-6-phosphate dehydrogenase, DIDS[(4,4-diisothiocyanatostilbene-2,2-disulfonic acid disodium salt hydrate],phosphoarginine and sodium metasilicate were obtained from Sigma-Aldrich(Taufkirchen, Germany).

Live specimens of Suberites domuncula Olivi (Porifera,Demospongiae, Hadromerida) were collected near Rovinj (Croatia) and then kept in aquaria in Mainz (Germany) for more than 2 years prior to their use.

The procedure for the formation of primmorphs (3D-cell cultures) from single cells was as described previously(Custodio et al., 1998; Müller et al., 1999). Starting from single cells, primmorphs of 3-7 mm are formed after 5 days. 7-day-old primmorphs were used for the experiments, cultivated in natural seawater supplemented with 0.2% of RPMI1640 medium. These 3D-cell cultures were transferred into RPMI1640 medium. Incubation was performed in the presence of the 2 μmol l^-1 silicic acid (ambient concentration,present in the natural seawater) or in seawater adjusted to a silicic acid concentration of 60 μmol l^-1(Krasko et al., 2002) and incubated for up to 5 days. Then the samples were used for the determination of arginine kinase activity and for in situ hybridization analysis. Where indicated, DIDS was added at a final concentration of 100 μmol l^-1; the incubation studies were performed in darkness. In parallel, samples were taken for extraction of RNA from liquid-nitrogen-pulverized sponge tissue using TRIzol Reagent (GibcoBRL, Grand Island, NY, USA), as recommended by the manufacturer.

These studies were performed with primmorphs incubated either for 5 days in the absence of additional silicic acid in the RPMI1640/seawater medium, or cultivated in RPMI1640/seawater supplemented with 60 μmol l^-1silicic acid.

The technique of differential display for identification of the arginine kinase cDNA was as described (Müller et al., 2002, 2003a). RNA (1 μg) from controls and from specimens treated with silicic acid was reverse-transcribed using the T₁₁CC oligonucleotide as 3′ primer. The resulting cDNA was added to the polymerase chain reaction (PCR) using the arbitrary primer GTGATCGCAG and the T₁₁CC primer in the assay, together with[α³²P]-dATP. After amplification the radioactive fragments were separated on a 6% polyacrylamide sequencing gel and autoradiographed. The major DNA bands were collected and subsequently reamplified in 50 μl reactions. Finally, the products were subcloned in pGEM-T vector(Promega, Madison, WI, USA) and sequenced. Among the over 50 sequences obtained, five partial cDNAs were isolated whose deduced polypeptide sequence showed similarity to that of arginine kinase. The complete cDNA was obtained by primer walking (Ausubel et al.,1995; Wiens et al.,1998). The arginine kinase cDNA was 1355 nucleotides (nt)long and was termed SDAK.

Sequences were analyzed using computer programs BLAST 2003(http://www.ncbi.nlm.nih.gov/blast/blast.cgi)and FASTA 2003(http://www.ncbi.nlm.nih.gov/BLAST/fasta.html). Multiple alignments were performed with CLUSTAL W Ver. 1.6(Thompson et al., 1994). Phylogenetic trees were constructed on the basis of amino acid (aa) sequence alignments by neighbour-joining, as implemented in the `Neighbor' program from the PHYLIP package (Felsenstein,1993). The distance matrices were calculated using the Dayhoff PAM matrix model as described (Dayhoff et al.,1978). The degree of support for internal branches was further assessed by bootstrapping (Felsenstein,1993). The graphic presentations of the alignments were prepared with GeneDoc (Nicholas and Nicholas,1997).

RNA was extracted from liquid-nitrogen-pulverized tissue using TRIzol Reagent (GibcoBRL) as described (Grebenjuk et al., 2002). Then 5 μg of total RNA was electrophoresed and blotted onto Hybond-N⁺ nylon membrane (Amersham; Little Chalfont,Bucks, UK). Hybridization was performed using a 350 nt segment of the SDAK cDNA; a 400 nt segment of the house-keeping gene β-tubulin of S. domuncula, SDTUB (EMBL/GenBank accession number AJ550806), was used as an internal standard. The probes were labeled using the PCR-digoxigenin (DIG)-Probe-Synthesis Kit (Roche, Mannheim, Germany). After washing, DIG-labeled nucleic acid was detected with anti-DIG Fab fragments and visualized by chemiluminescence technique using CDP (Roche). For semiquantitative analysis of the expression levels, the bands on the film were scanned using the GS-525 Molecular Imager (Bio-Rad, Hercules, CA, USA).

The method used was based on a described procedure(Polak and McGee, 1998; Perović et al., 2003). Frozen sections (8 μm) were prepared, fixed, treated with Proteinase K and subsequently fixed again with paraformaldehyde. To remove the sponge color the sections were washed with increasing concentrations of ethanol and decreasing concentrations of acetone. After rehydration the sections were hybridized with the labeled probe, a 250 nt long SDAK cDNA portion. After blocking,the sections were incubated with an anti-digoxigenin antibody conjugated with alkaline phosphatase. The dye reagent NBT/X-Phosphate was used for visualization of the signals. Antisense and sense ssDNA DIG-labeled probes were synthesized by PCR using the PCR-DIG-Probe-Synthesis Kit (Roche). Sense probes were used in parallel as negative controls in the experiments.

Primmorphs were exposed to silicic acid in the absence or presence of DIDS as indicated. Samples were homogenized in lysis-buffer [1× TBS(Tris-buffered saline), pH 7.5, 1 mmol l^-1 EDTA and protease inhibitor cocktail (1 tablet/10 ml; Roche)], centrifuged and the supernatants subjected to enzyme activity determination.

Arginine kinase activity was determined according to a modified procedure(Nealon and Herderson, 1976). The reaction mixture (final volume 1 ml) contained 10 mmol l^-1Tris-maleate buffer (pH 7.0) supplemented with 0.4 mmol l^-1phosphoarginine, 10 mmol l^-1 glucose, 5 mmol l^-1MgSO₄, 1 mmol l^-1 NADP⁺, 5 units ml^-1 hexokinase, 1 unit ml^-1 glucose-6-phosphate dehydrogenase and 10 μl of enzyme preparation. The appearance of NADPH was monitored in a spectrophotometer at 340 nm (25°C) and standardized using known amounts of ATP. Arginine kinase activity was expressed on the basis of protein content and given as nmol ATP min^-1 μg^-1protein. Five parallel experiments were performed.

Transmission electron-microscopic (TEM) observations were performed using a Zeiss (Aalen, Germany) EM 9A electron microscope. Primmorphs were pre-fixed in 2% glutaraldehyde in 0.05 mol l^-1 sodium cacodylate buffer (pH 7.4)with 0.2 mol l^-1 sucrose. After 3 days the specimens were post-fixed in 1% osmium tetroxide in 0.05 mol l^-1 cacodylate buffer(pH 7.4) at 4°C for 2 h. After dehydration the specimens were embedded in araldite; thin sections (700 Å) were double stained in uranyl acetate and lead citrate (Müller et al.,1986).

Protein concentration was measured according to Lowry et al.(1951) using bovine serum albumin as standard. The concentration of silicate in the natural seawater used for the experiments was determined, applying the molybdate (Silicon Test)method (Simpson et al.,1985).

The differential display of mRNA was examined in primmorphs grown in the absence or presence of silicic acid (60 μmol l^-1) for 5 days. RNA was isolated, reverse transcribed and the cDNA was used for PCR. The fragments were separated on sequencing gels. A comparison of the mRNA expression of non-treated versus silicic acid-treated primmorphs showed that among the >50 fragments sequenced, 10% encoded a putative arginine kinase. The complete sequence was obtained by primer walking. The arginine kinase cDNA, SDAK, had one open reading frame ranging from nt_89-91 to nt_{1229-1231 (stop)}. The 380 aa long deduced polypeptide, AK_SUBDO, has a putative size(M_r) of 43282 (Fig. 1A). Northern blot analysis performed with the sponge SDAK clone as a probe yielded one prominent band of ≈1.4 kb,indicating that the full-length cDNA had been isolated (see below).

The highest similarity of the S. domuncula AK_SUBDO was found with arginine kinase cDNA from the cnidarian (Anthozoa) Anthopleura japonica (Suzuki et al.,1997), with an `Expect value [E]'(Coligan et al., 2000) of 6e^-65 and an overall score of similar and identical aa residues of 51% and 35%, respectively. Therefore, the sponge sequence was termed arginine kinase. The sponge polypeptide displays the two characteristic domains of the ATP:guanido phosphotransferases (ISREC server; http://hits.isb-sib.ch/cgi-bin/PFSCAN_parser),the N-terminal (between aa₂₉ to aa₁₀₉; Pfam PF02807[http://www.sanger.ac.uk/cgi-bin/Pfam/getacc?])and C-terminal catalytic domains (aa₁₂₀ to aa₃₈₀; Pfam PF00217). These domains are implicated in the reversible transfer of high energy phosphate from ATP to the various phosphogens. The aa residues coordinating transfer of Mg²⁺ to ATP(Suzuki et al., 1997) are also present (Fig. 1A).

Arginine kinases, like all other members of the guanidino kinases, are restricted primarily to Metazoa; so no rooting could be performed when constructing a phylogenetic tree. The phylogenetic tree comprised the `usual'mono-domain 40-kDa guanidino kinases (arginine kinases) and the `unusual'two-domain 80 kDa guanidino kinases (arginine kinases). The two-domain kinases identified in the Cnidarian Anthopleura japonica(Suzuki et al., 1997) and the mollusks (Bivalvia) Solen strictus(Suzuki et al., 2002), Corbicula japonica (Suzuki et al., 2002) and Ensis directus(Compaan and Ellington, 2003),were dissected into two parts and both parts were aligned in parallel with the mono-domain guanidino kinases from mollusks, oyster Crassostrea gigasand Solen strictus, crustaceans Homarus gammarus(KARG_HOMGA, 538542), Eriocheir sinensis and Artemia franciscana, and insects, Drosophila melanogaster and Schistocerca americana. In addition, glycocyamine kinase from the annelid Nereis virens, human creatine kinase(Mariman et al., 1987) and the holothurian enzyme from Stichopus japonicus were included in the alignment. The tree shows very strikingly that the duplication of the arginine kinases occurred in some mollusk species (e.g. S. strictus), but not in others (e.g. C. gigas; Fig. 1B). It had been proposed recently(Takeuchi et al., 2004) that gene duplication and their subsequent fusion in mollusks occurred separately from events that took place in hydrozoa (A. japonica).

In order to clarify whether the S. domuncula arginine kinase gene is upregulated after incubation with silicic acid, primmorphs were cultured in the presence or absence of 60 μmol l^-1 additional silicic acid. Northern blot analyses were performed with the labeled cDNA probe for arginine kinase (SDAK) and the house-keeping gene β-tubulin(SDTUB). These studies revealed that after addition of silicic acid to the primmorphs, within 1 day a significant (threefold) increase of the steady-state level of arginine kinase transcripts could already be measured, a value which further increased after additional 2-4 days(Fig. 2, left). In contrast, if the primmorphs remained in culture without additional silicic acid, no significant change of the expression level could be seen(Fig. 2, right). Control blots,performed with the house-keeping gene β-tubulin, confirmed that the same amount of RNA was loaded onto the gels. In parallel, expression studies were performed in the presence of 100 μmol l^-1 DIDS for 0-5 days, and in the presence of this inhibitor the silicic acid-mediated upregulation of arginine kinase expression was not seen (data not shown).

As further proof that in primmorphs gene expression of arginine kinase is positively affected by silicic acid, in situ hybridization studies were performed. Primmorphs were incubated for 5 days in the absence(Fig. 3A,B) or presence of 60μmol l^-1 of silicic acid(Fig. 3C,D). Then cryosections were hybridization with the labeled probe for arginine kinase (SDAK). Sections from primmorphs that remained for 5 days in the absence of silicic acid did not show high signals using antisense SDAK ssDNA probes(Fig. 3A,B). However, if the primmorphs had been kept in the presence of silicic acid, high expression of this gene was seen, especially in the center of these 3D-cell cultures(Fig. 3C,D). In a further series, the primmorphs were cultivated in the presence of silicic acid together with the inhibitor DIDS (100 μmol l^-1; Fig. 3E,F). In sections from these primmorphs the expression level was almost identical to that seen in cultures without additional silicic acid. No reaction was observed when the cells were treated with sense probes (not shown).

Primmorphs were obtained as described under `Materials and methods'. After formation of the 3D-cell cultures, cultivation in seawater was continued in the absence of additional silicic acid for 5 days. Under these conditions,arginine kinase activity was approximately 2.5 nmol ATP generated min^-1 10 μg^-1 protein. This value did not change significantly when primmorphs were incubated with 100 μmol l^-1DIDS (Fig. 4). However, if the cultures were incubated with an additional 60 μmol l^-1 of silicic acid the enzyme activity had already increased after 1 day to 6 nmol ATP min^-1 10 μg^-1 protein, a value which rose to 27 nmol ATP min^-1 10 μg^-1 protein on further incubation. Coincubation with DIDS reduced the enzyme activity by more than 70%(Fig. 4). These data show that the arginine kinase activity increased strongly in primmorphs in response to silicic acid.

We have previously demonstrated that primmorphs start to form spicules after incubation with 60 μmol l^-1 silicic acid(Krasko et al., 2000). Here we cut cross sections through primmorphs that had been cultivated for 5 days in the absence or presence of 60 μmol l^-1 silicic acid. TEM analysis of primmorphs grown in the absence of silicic acid did not show any cells that synthesized spicules. However, in primmorphs kept in the presence of silicic acid, sclerocytes that contained newly growing spicules were frequently observed (Fig. 5). In Fig. 5A two sclerocytes are seen, each surrounding a spicule. At higher magnification it can be observed that the blunt end of one spicule is tightly associated with two fibrils,which are very likely collagen fibers (Fig. 5B,D); however such fibers were not always present at the blunt end (Fig. 5C). Those fibers surrounded the spicules at all phases (Fig. 5B-D). Note that the sharp, likewise growing, tip of the spicule showed an inhomogeneous structure, suggesting that the material in this regions is not densely packed (Fig. 5D).

The phylum Porifera comprises species of sizes between 2-4 mm and up to 2 m; their bodies are structured into organ-like tissues, which are not connected through a nervous coordination network (see Müller et al., 2004b). Even contraction of channels and pumping of the water through the aqueous system are not mediated by muscle organs. Muscles evolved first in the second diploblastic phylum, the Coelenterata. Therefore it might be expected that in sponges phosphagen (creatine), which is the characteristic energy source in muscles for the generation of ATP, is not present, nor the corresponding energy transferring enzyme, creatine kinase. The latter assumption is supported by database searches in the S. domuncula EST library, which does not contain a cDNA for this enzyme. However, in invertebrates other phosphorylated high-energy guanidine phosphagens, e.g. phospho-arginine or phospho-glycocyamine, are known to serve as energy donors for the formation of ATP in the cytoplasm. As in most cells, in sponges ATP is also generated by oxidative phosphorylation (see Sona et al., 2004). The phosphagen kinases are localized at the sites of ATP hydrolysis and/or synthesis, allowing a temporal and spatial ATP buffer system as well as a tuned regulation of the intracellular phosphate concentration (reviewed in Ellington,2001). In the present study the arginine kinase activity in S. domuncula was determined to be approximately 2.5 nmol ATP generated min^-1 10 μg^-1 protein, a value that is within the range of arginine and creatine kinase activities measured in the primitive-type sperm of certain marine invertebrates(Tombes and Shapiro,1989).

The technique of differential display of mRNA was used to identify which energy transferring kinase is involved in one major energy-consuming reaction in sponges, the formation of the skeleton. Primmorphs were exposed to silicic acid and their mRNA expression patterns were compared with those in 3D-cell aggregates, which grew in the low-silicic acid medium. Silicic acid is the inorganic component of the spicules in Demospongiae. As outlined in the Introduction, the spicules are the major skeletal elements of sponges; their framework is connected by filamentous organic fibrils(Garrone, 1978). Furthermore,it should be mentioned that the growth of spicules is a fast process(Elvin, 1972). The synthesis and formation of this scaffold require considerable energy reserves. Analytical studies revealed that arginine is one dominant amino acid in sponges, reaching intracellular concentrations of 2 mmol l^-1 (Marco Giovine; unpublished observations). The differential display analyses revealed that 10% of the transcript fragments contribute to the arginine kinase cDNA. The full-length cDNA was identified. It comprises, as with all other metazoan phosphagen kinases, the two domains characteristic for these kinases, the N-terminal domain and the C-terminal catalytic domain. Also the Mg²⁺/ATP coordinating residues within the C-terminal domain are conserved.

To study the phylogenetic aspect of the sponge arginine kinase was equally interesting. In yeast (Saccharomyces cerevisiae) or plants (e.g. Arabidopsis thaliana) no arginine kinases, or related enzymes, exist. Hence the sponge molecule is evolutionarily one of the oldest phosphagen kinases. This fact is also reflected by the radial phylogenetic tree, which shows that the human/vertebrate creatine kinases, and the glycocyamine kinase(Nereis virens) branched off from the sponge common ancestor molecule(Fig. 1B). Arginine kinases emerged in the other direction, comprising the `usual' mono-domain 40-kDa guanidino kinases (arginine kinases) and the `unusual' two-domain 80 kDa guanidino kinases (arginine kinases). Interestingly, the next closest arginine kinase family member, the cnidarian A. japonica(Suzuki et al., 2002; Takeuchi et al., 2004) has already undergone an independent gene duplication in another taxon, the mollusks (see Fig. 1). While the guanidino kinase from the mollusk C. gigas belongs to the mono-domain kinases, other molluskan kinases, from S. strictus, C. japonica and E. directus, are two-domain 80 kDa guanidino kinases. It is amazing that within the Bivalvia, and even in the same subclass Lammellibranchia, this duplication proceeded from mono-domain, C. gigas (superorder Filibranchia), to two-domain kinases, S. strictus (Eulamellibranchia), C. japonica (Eulamellibranchia)and E. directus (Eulamellibranchia). It was suspected that this gene duplication parallels the higher enzyme activity(Takeuchi et al., 2004). Crystal structure classification (Cheek et al., 2002) supports the grouping of the kinases deduced from alignment studies with protein sequence data. The enzyme activity of arginine kinase has been detected in the flagellate protozoa Trypanosoma cruzi(Alonso et al., 2001) and Paramecium caudatum (Noguchi et al., 2001), indicating that arginine kinases evolved before the emergence of multicellular animals. Recent phylogenetic analyses also indicates that the ciliate arginine kinases show a high phylogenetic relationship to those of sea anemones(Ellington and Suzuki,2005).

The primmorph system was established in order to obtain molecular biological insights into the proliferation(Le Pennec et al., 2003),differentiation (Müller et al.,2004b) and immunological(Müller et al., 2002)capacities of primordial sponge cells. Here, this system was used to demonstrate that the expression level of the gene encoding the arginine kinase strongly depends on the presence of silicic acid in the medium. The gene induction is fast; after 1 day a strong upregulation is already seen. This fast response is characteristic for the primmorph system. Previously it was shown that the genes for silicatein, the enzyme that causes the formation of silica in the spicules, and for collagen, the organic cover around the skeletal elements, also already show strong induction 1 day after silicic acid incubation (Krasko et al.,2000). One piece of evidence that the increased expression of the arginine kinase gene is indeed dependent on the exogenous silicic acid, came from inhibition studies using DIDS, an established inhibitor of band 3-mediated anion exchanger through covalent binding to lysine residue(s)(Kopito, 1990). The inhibition is not restricted to the transport of the

One major differentiation pathway is the transition from the totipotent stem cells, the archaeocytes, into the spicule-forming sclerocytes(Müller et al., 2003b). Archaeocytes are the major cell fraction in primmorphs(Müller et al., 1999) and the inducer of this differentiation direction is silicic acid; in the archaeocyte cell lineage noggin and MSCP expression are crucial for the fate of the sclerocyte differentiation(Schröder et al., 2004b). In the present study it is seen that the expression of the arginine kinase gene in primmorphs, incubated in the presence of silicic acid,occurs primarily in the center of these 3D-cell aggregates. This localized expression is in accordance with previous findings that differentiation of cells, e.g. for the formation of aqueous canals, is first detected in the center of the primmorphs (Müller et al., 2004a). It interesting to observe this specific localization in the choanocyte-forming region of primmorphs. In fact, these peculiar cells need high amounts of energy for cilia movement and phosphoarginine should be primarily involved as energy reservoir. It is intriguing to notice that in ciliates (Protozoa) arginine kinase activity also plays a fundamental role in cilia movement (Noguchi et al.,2001). The specificity of the silicic acid-mediated induction of the arginine kinase gene is, again, shown by inhibition studies with DIDS; this stilbene analogue almost completely prevents this expression.

The increased expression of arginine kinase also correlates with a strong enhancement of arginine kinase enzyme activity. In primmorphs a strong increase of activity was measured after 1 day of exposure to silicic acid,again a process that could be blocked by DIDS. Since this inhibitor blocks both spicule formation and expression of arginine kinase gene, it can be assumed that under these conditions a steep electrochemical gradient for silicic acid transport results. An association of phosphagen kinases with membrane-associated ATPase is well documented (reviewed in Ellington, 2001). Final support for the differentiation-mediating capacity of silicic acid, which allows formulation of a chain of reactions from increased expression of the arginine kinase gene to enhanced enzyme activity and finally to the onset of spicule formation in primmorphs, was documented by transmission electron-microscopy. This analysis revealed that spicules are formed in those primmorphs that were exposed to silicic acid.

Taken together, the results of the present work show that arginine kinase is a crucial enzyme in the high energy-consuming reaction of spicule formation in primmorphs. In addition, the data demonstrate that the evolutionary oldest metazoan phylum, the Porifera, comprise the ancestral phosphagen kinase, an arginine kinase, for all other kinases of higher taxa. Studies are in progress to elucidate the role of arginine kinase in adult specimens; there the highest energy consumption should be localized in the choanocytes, which are the`motor' cells that drive the water through the aqueous canal system. A similar role for creatine kinase in choanocyte function has recently been suggested for another sponge (Sona et al.,2004).

This work was supported by grants from the Deutsche Forschungsgemeinschaft,the Bundesministerium für Bildung und Forschung Germany (project: Center of Excellence BIOTECmarin) and the International Human Frontier Science Program (RG-333/96-M). Note that the cDNA sequence for arginine kinase from Suberites domuncula has been deposited (EMBL/GenBank) under the accession number AJ744770 (AK_SUBDO).