Sponges are filter feeders that consume a large amount of energy to allow a controlled filtration of water through their aquiferous canal systems. It has been shown that primmorphs, three-dimensional cell aggregates prepared from the demosponge Suberites domuncula and cultured in vitro, change their morphology depending on the light supply. Upon exposure to light, primmorphs show a faster and stronger increase in DNA, protein and glycogen content compared with primmorphs that remain in the dark. The sponge genome contains nocturnin, a light/dark-controlled clock gene, the protein of which shares a high sequence similarity with the related molecule of higher metazoans. The sponge nocturnin protein was found showing a poly(A)-specific 3′-exoribonuclease activity. In addition, the cDNA of the glycogenin gene was identified for subsequent expression studies. Antibodies against nocturnin were raised and used in parallel with the cDNA to determine the regional expression of nocturnin in intact sponge specimens; the highest expression of nocturnin was seen in the epithelial layer around the aquiferous canals. Quantitative PCR analyses revealed that primmorphs after transfer from light to dark show a 10-fold increased expression in the nocturnin gene. In contrast, the expression level of glycogenin decreases in the dark by 3– 4-fold. Exposure of primmorphs to light causes a decrease in nocturnin transcripts and a concurrent increase in glycogenin transcripts. It was concluded that sponges are provided with the molecular circadian clock protein nocturnin that is highly expressed in the dark where it controls the stability of a key metabolic enzyme, glycogenin.
In higher metazoans, the 24 h periodicity in the environment contributes to the evolution of the molecular circadian clock . In mammals the hypothalamic region, the suprachiasmatic nucleus, was found to harbour the master pacemaker of circadian rhythm . Moreover, peripheral organs have recently also been shown to contain circadian oscillators . The generation of rhythms in the hypothalamus is assumed to be based upon molecular feedback loops within a positive limb comprising the helix–loop–helix transcription factors CLOCK (circadian locomotor output cycles kaput) and BMAL1 [brain and muscle ARNT (aryl hydrocarbon receptor nuclear translocator)-like 1], and a negative oscillator limb with two Per (period) and two CRY (cryptochrome) proteins that are interconnected via the nuclear orphan receptor REV-ERBα (reviewed in ). Light signals are required for clock synchronization of the suprachiasmatic nucleus; those photons are perceived by rhodopsin photoreceptors in the outer retina and also by melanopsin-containing ganglion cells in the inner retina. The electric signals, originating from those photic inputs, are transmitted as electrical signals via neurons to the retino-hypothalamic tract via a cascade of neurotransmitters, allowing the influx of Ca2+ into cells of the suprachiasmatic nucleus after which differential gene expression occurs. The hierarchy of the control and expression levels of the molecular oscillator has been elucidated to a large extent (see ).
In vertebrates, the CLOCK/BMAL1 transcription factors integrate metabolic systems both directly and indirectly and control the energy balance and metabolism at multiple levels within the CNS (central nervous system) and in the peripheral tissues . The homoeostasis of the protein and lipid composition in animals is under the control of the circadian clock (see ). In turn, there is an increasing awareness that a series of cellular genes are expressed in a circadian rhythm. The circadian-responsive system acts both on the transcriptional level, with CLOCK and BMAL1 as the major factors (see ), and also on the post-transcriptional level, e.g. nocturnin [6,7]. Nocturnin, a light-responsive circadian regulator, is under the control of the CLOCK/BMAL1 transcriptional system .
The circadian clock is not restricted to higher metazoans, it is also present in protostomians and deuterostomians and has also been described in corals . Therefore it is also reasonable to propose such a mechanism for the lowest metazoans, the sponges (phylum Porifera). It had been described that both adult sponges and their larvae perceive light , suggesting that they might be capable of adapting to the environmental light/dark cycles. However, until now no photopigments have been described from sponges that might be able to absorb light and in turn undergo chemical changes. Recently, retinoids had been described to exist abundantly in the demosponge Suberites domuncula ; however, a function, as for retinal in the rhodopsin system, could not been established. This was expected since an opto-visual sense apparently does not exist in sponges; however, photic inputs are received and generated by sponges. The larvae of the demosponge Reneira sp. react to diffused or focused external light . Some species of sponges, e.g. S. domuncula, have the ability to generate light via the enzyme luciferase [12,13], suggesting that the photons, generated enzymatically, are transmitted via the biosilica glass skeleton to the FAD-containing cryptochrome that might act as a photosensor . The existing data led us to suggest that this proposed photoreception/phototransduction circuit functions as a nerve-cell like signal transmitting system. Interestingly enough, S. domuncula reacts to different light wavelengths with a differential gene expression of, for example, luciferase, CRY and the transcription factor SOX [SRY (sex-determining region Y)-box] .
Sponges, as filter-feeders, are decisively dependent on the nutritional status of the environment. From studies on vertebrates, it is known that one of the circadian expressed proteins, nocturnin, is induced during fasting in the white adipose tissue of mice  and plays a role in linking nutrient sensing by the circadian clock primarily to lipid mobilization in the fat cells. Nocturnin had been found to be a deadenylase, a 3′-exoribonuclease, with specificity for poly(A) tracts in mRNA, whose steady-state expression is under both circadian and acute regulation . The poly(A)-specific 3′-exoribonuclease had been initially discovered and purified from calf thymus by us [16–18] and had been assigned the EC number 22.214.171.124 [poly(A)-specific ribonuclease; see http://www.brenda-enzymes.org/php/result_flat.php4?ecno=126.96.36.199]. Subsequently, the 3′-exoribonuclease was found to control the half-life of mRNAs through hydrolysis of their poly(A) tracts . This nocturnin enzyme is expressed with high amplitude rhythms in peripheral tissues in mammals, with peak levels at night [6,7]. Until now only a few target mRNA species had been identified upon which also nocturnin acts .
Focusing on sponges, we decided to identify and subsequently to clone the cDNA encoding nocturnin in order to get the first molecular insight into the (potential) circadian rhythm of this enzyme and to correlate its different circadian levels with the intermediary metabolism of these animals, using S. domuncula as an experimental model. We have selected a key enzyme from carbohydrate/glycogen metabolism, glycogenin, as a target protein whose circadian expression level has been determined in parallel to nocturnin. Glycogenin is an enzyme that initiates glycogen synthesis, and whose expression is likely to be under the control of nocturnin in mammals . It functions as a primer for glycogen synthesis, allowing the polymerization of the first few glucose units on its enzyme molecule, after which other enzymes take over the further elongation of the glycogen chains .
MATERIALS AND METHODS
Chemicals, materials and enzymes
The sources of the materials used in the present study were described previously .
Sponges and primmorphs
Specimens of the marine sponge S. domuncula (Porifera, Demospongiae, Hadromerida) were sampled in the Northern Adriatic near Rovinj (Croatia), and then kept in aquaria in Mainz (Germany) at a temperature of 17°C for more than 12 months .
Primmorphs, three-dimensionally growing cell aggregates in vitro, were prepared from single cells, obtained from S. domuncula, as described previously . They were cultivated for up to 10 days in natural seawater (Sigma) supplemented with organic nutrients (0.2% RPMI 1640 medium; Gibco). After 1 day in culture, the cells form 100–200 μm aggregates which increased in size during the following 5 days. Unless otherwise stated, 3-day-old aggregates were used. Where indicated, the primmorphs remained in complete darkness or had been illuminated with 30 lux, using a 15-watt Cool White fluorescent light (Westinghouse Electric) and kept in a temperature-controlled incubator at 16°C.
In one series of experiments, intact animals were exposed to 200 lux by a white fluorescent light penetrating through the 5–10 cm of water in front of the animals. Prior to analysis by immunohistology, those animals remained under continuous light exposure or in complete dark for 3 days.
Identification and cloning of the poriferan nocturnin and glycogenin cDNAs
The two cDNA sequences [SDNOC (S. domuncula nocturnin)] and [SDGYG (S. domuncula glycogenin)] were isolated. Details are given in the Supplementary Online data (at http://www.BiochemJ.org/bj/448/bj4480233add.htm).
Preparation of recombinant nocturnin and antibodies
The recombinant nocturnin (rNOC_SUBDO) was prepared in Escherichia coli cells as described in the Supplementary Online data. This protein sample was used for raising polyclonal antibodies in rabbits.
SDS/PAGE and Western blot analysis
Details of the SDS/PAGE and for Western blot analyses are given in the Supplementary Online data.
qRT–PCR (quantitative real-time reverse transcription PCR)
The absolute steady-state expression level of SDNOC, as well as of SDGYG, was determined by qRT–PCR. From single cell suspensions, where primmorphs were formed, RNA was isolated, as described with the respective experiments in the Results section.
The details of the qRT–PCR procedure to determine the expression level of SDNOC were described previously [13,23]. All reactions were run with an initial denaturation at 95°C for 10 min, followed by 40 cycles each at 95°C for 20 s, then 60°C for 20 s and 72°C for 35 s. The following primers were used for amplification of SDNOC: forward, 5′-GCCTT-GTTCTTTCGCTCATC-3′ (nt 413–432) and reverse, 5′-AGACTTCTTCAGTCCCAACC-3′ (nt 535–516); PCR product size is 123 bp. As a reference housekeeping gene, we used SDGAPDH (S. domuncula glycerol 3-phosphate dehydrogenase; GenBank® accession number AM902265.1] with the following primers: forward, 5′-TCCAAACCAGCCAAGTACGATG-3′ (nt 816–837) and reverse, 5′-AGTGAGTGTCTCCCCTGAAGTC-3′ (nt 945–924); fragment size was 130 bp. After qRT–PCR, the threshold position was set to 50.0 RFU (relative fluorescence units) above PCR-subtracted baseline for all runs. Mean Ct values and efficiencies were calculated by the iCycler software. Relative expression level was calculated by EGAPDHCtGAPDH/ENOCCtNOC, where E is the PCR efficiency and Ct is the threshold cycle.
The quantitative steady-state expression level of SDGYG was determined in parallel assays to those of nocturnin expression. For this series of experiments the following primers were used: forward, 5′-GGCAACATCACTATGTTCCAGAGG-3′ (nt 732–755) and reverse, 5′-ATCACTCTGTCCGATCTGCTCG-3′ (nt 868–847); PCR product size of 137 bp. The expression level was calculated by EGAPDHCtGAPDH/EGYGCtGYG.
Tissue slices were prepared as described previously . Slices of 8 μm thickness were fixed in paraformaldehyde. The specimens were incubated with the primary antibodies, the rabbit anti-nocturnin serum (PoAb-rNOCSUBDO), diluted 1:1500 in blocking solution and incubated while shaking at 4°C overnight. As controls, both pre-immune serum and PoAb-rNOCSUBDO antibodies, adsorbed with recombinant nocturnin, were used. Unbound antibodies were removed by washing four times with PBS prior to the incubation with fluorescently labelled [with Cy5 (indodicarbocyanine)] secondary antibodies (1:3000 dilution). Parallel slices were stained with DAPI (4′,6-diamidino-2-phenylindole; Sigma). The slices were inspected with an Olympus AHBT3 microscope under immunofluorescence light at an excitation light wavelengths suitable for either Cy5-stained structures or for DAPI.
Determination of the glycogen level
The microdetermination method, performed as described by Templeton , was used to quantify the concentration of glycogen in the primmorphs. In brief, the anthrone reagent (catalogue number 10740, Sigma; for determination of carbohydrates) was used, which was applied in the assay after washing the sample in distilled water and alcohol and subsequent drying as described previously . Glycogen was extracted and precipitated from hydrolysates with 80% ethanol and treated with ethanol to remove low-molecular-mass carbohydrates. The amount of glycogen in the samples was determined using the glycogen from Mytilus edulis (G1767, Sigma) as a standard. Optical density was determined against a water blank at 630 nm in a spectrophotometer. Protein determinations were performed in homogenates prior to centrifugation.
In order to prove that nocturnin has enzymatic activity as a 3′-exoribonuclease, the full-length nocturnin protein (rNOCPAE_SUBDO) was prepared recombinantly (details are given in the Supplementary Online data). The 37.5 kDa protein obtained was assayed for deadenylase activity using the poly(A) tail of a partial glycogenin cDNA. A part of the SDGYG cDNA comprising the ORF (open reading frame) nucleotide sequence (from nt 465–575) was cloned into the pSP64 poly(A) Vector system (Promega). This vector has a stretch of 30 dA:dT residues allowing the synthesis of a RNA containing a 30-residues-long poly(A) tail . To generate the synthetic poly(A)+ mRNA substrate, the glycogenin cDNA (nt 465–575) was digested with HindIII and SacI enzymes (New England Biolabs) and ligated into the vector, digested with the same restriction enzymes. The pSP64[poly(A)+] construct was transcribed using SP6 polymerase and the transcribed poly(A)+ mRNA was purified by phenol/chloroform extraction, precipitated and then resuspended in diethylpyrocarbonate-treated water . The mRNA concentration was measured at A260 and the transcript size was determined by gel electrophoresis followed by staining with SYBR Green II (Molecular Probes) [28,29]. RNA size markers (peqGOLD, PEQLAB) were used.
The deadenylase assay was performed as described previously [16,17,30] and contained, in a total reaction volume of 80 μl, the following components: 100 mM Tris/HCI (pH 8.7), 2 mM MgCl2, 0.2 mM dithiothreitol, 0.02% Nonidet P40, 0.3% spermidine and 10 pg/ml BSA, as well as 10 μg of poly(A)+ glycogenin mRNA. The enzyme reaction was started with either 3 μg of recombinant rNOCPAE_SUBDO or 1 μg of poly(A)-specific 3′-exoribonuclease, purified from the calf thymus . After incubation at 37°C, aliquots of 20 μl were size-separated, stained with SYBR Green II and the bands were visualized with a short-wave ultraviolet transilluminator .
Northern blot analysis
RNA was isolated from primmorphs, essentially as described previously . RNA was extracted from liquid-nitrogen-pulverized tissue with TRIzol® Reagent (Invitrogen). The concentration and the purity of the RNA were determined optically using the absorbance at 260 nm and the A260/A280 ratio respectively.
Aliquots of total RNA (5 μg) were electrophoresed through 1% formaldehyde/agarose gel and blotted on to Hybond-N+ nylon membranes following the manufacturer's instructions (Amersham) as described previously . Hybridization was performed with a 0.37 kb part of the SDGYG cDNA, spanning the region nt 13–376 within the coding region. The probe was labelled with the PCR DIG (digoxigenin) probe synthesis kit according to the manufacturer's instructions (Roche). After washing, DIG-labelled nucleic acid was detected with anti-DIG Fab fragments (conjugated to alkaline phosphatase; 1:10000 dilution) and visualized by a chemiluminescence technique using CDP-Star, according to the instructions of the manufacturer (Roche) and as described previously . The ssRNA (single-stranded RNA) ladder from New England BioLabs was used as a marker for the size determination.
The poly(A) tails of the mRNAs were digested with RNAseH as described previously . In brief, 50 μg of RNA was denatured by heating and immediately chilling on ice. Then the samples were incubated at 25°C for 10 min either in the absence or in the presence of oligo(dT)12–18 (Invitrogen). The hybrid RNA:oligo(dT) samples were incubated (37°C for 30 min) with 3 units of RNase H (Invitrogen) per 50 μl assay. Those digested RNA samples were then used for Northern blot analysis.
For the quantification of protein, the Bradford method (Roti-Quant solution, Roth)  was used. DNA content was measured as described previously . In order to determine the dry mass, primmorph samples were dried at 105°C and then weighed on a microbalance (Sartorius Expert LE26P). Six parallel determinations were performed.
S. domuncula and its primmorphs
Three different colour types of S. domuncula are found in nature , primarily as red specimens (Figure 1A), secondarily as yellow ones (Figure 1A) and finally as whitish/magenta ones. The specimens live in association with the hermit crab Pagurites oculatus (Decapoda, Paguridea; Figure 1A) which resides inside of shells of the mollusc Trunculariopsis trunculus (Gastropoda, Muricidae). For the experiments described in the present study, the common red specimens were used. On the surface of both colour types, gastropods/molluscs Bittium sp. (Mollusca, Gastropoda, Cerithiidae, Bittium) occasionally graze.
Specimens of S. domuncula and primmorphs, the three-dimensional aggregates obtained from them
Effect of light on morphology and DNA, protein and glycogen content in primmorphs
If exposed to the light, the small aggregates, formed during the first 2 days, re-aggregate to larger units with sizes of around 1 mm at day 4 (Figures 1B and 1C). In contrast, those three-dimensional cell aggregates which were left in the dark showed a significantly retarded re-aggregation; only small aggregates are seen at day 2 and the primmorphs at day 4 show a lobated morphology and have not yet started to form round-shaped units (Figures 1D and 1E).
The DNA and protein content in growing primmorphs differs in dependence on the light conditions used for cultivating them. During continuous illumination of the primmorphs, 1 day after the onset of the re-aggregation process the DNA and protein content was determined as 0.95 and 125 μg/mg of dry mass respectively. During the following 4 days, the mass increased to 2.9 and 211 μg/mg of dry mass of DNA and protein respectively (Figure 2). When kept in the dark, the primmorphs show a similar increase in DNA content during the incubation period of 0–4 days; however, the content of protein in the primmorphs is significantly lower, with a value of 105 μg/mg of dry mass at day 0 and 142 μg/mg of dry mass at day 4.
DNA and protein content in primmorphs grown in the light or dark
The content of glycogen showed similar kinetics. Under light exposure, the glycogen content in the primmorphs increased significantly from 28 μg/mg of dry mass to 52 and 67 μg/mg of dry mass at days 3 and 4 respectively. No significant increase was seen in the primmorphs kept in the dark; values around 15–22 μg/mg of dry mass are measured during the incubation period of 0–4 days (Figure 3).
Alteration of glycogen content in the primmorphs during light exposure or dark maintenance from 0–4 days
Sequence analysis of sponge nocturnin
The complete SDNOC cDNA was isolated (Supplementary Figure S1 at http://www.BiochemJ.org/bj/448/bj4480233add.htm). The size of the deduced protein (NOC_SUBDO), comprising 314 aa, is 34836 Da. As shown in the Supplementary Online data, the sponge nocturnin mRNA had a complete size of 1.2 kb. The deduced protein comprises the domain that is characteristic for the superfamily of exonuclease-endonuclease-phosphatases. In addition, a phylogenetic analysis revealed that the SDNOC polypeptide shared the highest sequence similarity with the corresponding cnidarian protein, Nematostella vectensis (NCBI reference sequence XP_001626780), comprising a similarity/homology score of 50%/35%.
Recombinant SDNOC and antibodies raised against it
A part of the cDNA comprising the ORF for nocturnin (NOC_SUBDO) was expressed in E. coli cells. After purification by affinity chromatography, the 30 kDa protein had a purity of >95% (see Supplementary Online data). This protein sample (rNOC_SUBDO) was used to raise polyclonal antibodies (PoAb-rNOCSUBDO). The serum recognized the 30-kDa recombinant protein after size separation by SDS/PAGE and Western blotting (Supplementary Figure S2, lane d at http://www.BiochemJ.org/bj/448/bj4480233add.htm). This protein did not cross-react with a pre-immune serum (Supplementary Figure S2, lane e).
Sequence analysis of SDGYG
The ORF of the SDGYG cDNA encodes a protein (GLYG_SUBDO) of a calculated size of 37023 Da; (Supplementary Figure S3 at http://www.BiochemJ.org/bj/448/bj4480233add.htm). In this sequence the characteristic domains for the glycogenins, the N-terminal β-α-β Rossmann-like fold as well as the glycosyl transferase family 8 segment, are present.
Expression of nocturnin in the tissue of the animals
The regional expression of nocturnin within the sponge tissue was analysed by immunofluorescence. Intact animals were kept for 3 days in either dark or under continuous light. Tissue slices within the surface region of a dark-kept animal were prepared and reacted with the anti-nocturnin antibodies (PoAb-rNOCSUBDO). The immunofluorescence image convincingly shows that the cells around aquiferous canals (the endo-pinacoderm layer) are brightly stained (Figures 4A and 4C). The cell arrangement around the canals was visualized by DAPI staining (Figures 4B and 4D). In contrast, when the slices of the light-maintained animals were reacted with PoAb-rNOCSUBDO almost no staining was seen (Figures 4E and 4G). In parallel images, the arrangements of the cells around the canals are highlighted by DAPI staining (Figures 4F and 4H). Two control series were performed to demonstrate that the immunofluorescence signals are specific. First, when slices from a dark-kept animal were reacted with pre-immune serum, almost no staining is seen (Figure 4I). The corresponding DAPI-stained slice shows that a region around a canal was selected for sectioning (Figure 4J). Secondly, when adsorbed PoAb-rNOCSUBDO antibodies had been applied for the immune reaction, again no immunofluorescent signals were seen on the sections obtained from a dark-kept animal (Figure 4K); the corresponding DAPI image is shown in Figure 4(L).
Expression of nocturnin in tissue of S. domuncula as analysed by immunohistology and applying antibodies against nocturnin
Expression levels of nocturnin and glycogenin in primmorphs in either the light or dark
A quantitative determination of the steady-state expression level both for the SDNOC and the SDGYG was performed by using qRT–PCR. Primmorphs cultivated for 3 days under light exposure showed an expression level for SDNOC of 2.7×10−3, with respect to the steady-state expression of the SDGAPDH gene (Table 1). However, if those three-dimensional aggregates were kept for the same period of time in dark, the expression level significantly increased to 48.2×10−3. Contrasting day/night expression was also measured for SDGYG. The amount of transcripts is considerably higher in light-exposed primmorphs with 5.2×10−1, compared with the dark-maintained primmorphs with 1.2×10−1 (Table 1).
|Gene expression (mRNAx/mRNAGAPDH)±S.D.|
|Gene expression (mRNAx/mRNAGAPDH)±S.D.|
The time-dependent induction of SDNOC mRNA transcript formation during dark onset was determined in primmorphs that had been kept for 3 days under continuous light. After onset of complete darkness, the SDNOC transcript level increased significantly from 3.2×10−3 to 8.3×10−3, even after 3 h, and reached a maximum value of 49.0×10−3 after a 12-h incubation period (Figure 5). The decay kinetics of SDGYG expression proceeds in the opposite direction, starting with 475×10−3 transcripts, with respect to the SDGAPDH level, and reached a minimum after the 12–24 h incubation period with 153×10−3 and 138×10−3 transcripts respectively.
Reverse kinetics of the increase/decrease of the abundance of the nocturnin compared with glycogenin transcripts after transfer of primmorphs from light to dark
Similarly fast is the decay/synthesis kinetics of the SDNOC and SDGYG transcripts in primmorphs kept in the dark after onset of light exposure (Figure 6). The primmorphs were cultured in the dark during the first 3 days after starting aggregation and they were then exposed to light. Starting with the high transcript level for SDNOC of 41.7×10−3 with respect to the SDGAPDH transcripts, the decay starts after 3 h and significantly dropped to 29.6×10−3, after just a 3-h dark period, to almost the lowest level that is reached after 6–24 h of light exposure of approximately 10×10−3. The increase of the SDGYG transcript level becomes, again, significant even after a 3-h exposure to light (193×10−3) and reaches maximal levels after 12–24 h of 487×10−3 and 541×10−3 respectively.
Dark/light reciprocal expression studies of the nocturnin and the glycogenin gene; quantification through qRT–PCR
Synthetic SDGYG mRNA as a target for the deadenylase SDNOC
The SDGYG mRNA was synthesized using the pSP64 poly(A) Vector system. The poly(A)+ mRNA contained a 30-nt long poly(A) stretch and a 100-nt long partial SDGYG mRNA section. This mRNA was incubated with the full-length recombinant nocturnin protein from S. domuncula (rNOCPAE_SUBDO) or with the native poly(A)-specific 3-exoribonuclease, purified from the calf thymus . The incubation was performed at 37°C. After a 10-min incubation period, the poly(A)+ mRNA already became partially deadenylated, if assayed with the recombinant nocturnin (Figure 7, left-hand panel). After an extended incubation period for 30 min, the poly(A)+ mRNA added to the assay underwent complete deadenylation (Figure 7, left-hand panel), as did the native poly(A)-specific 3′-exoribonuclease (Figure 7, right-hand panel). This deadenylation of the poly(A)+ mRNA was prevented if the assays were supplemented with 10 mM EDTA, an inhibitor of the nuclease activity  (results not shown).
Nocturnin-mediated deadenylation of the synthetic poly(A)-containing mRNA substrate
Northern blot analysis of SDGYG mRNA from light- and dark-exposed primmorphs
The size of the mRNA encoding SDGYG was determined by Northern blot hybridization using the labelled SDGYG cDNA as a probe. The experiments revealed that the mRNA for SDGYG, isolated from primmorphs exposed to light had a larger size, compared with the transcripts isolated from primmorphs kept in the dark (Figure 8A). Correlating with the markers that run in parallel, the SDGYG mRNA existing in primmorphs kept for 96 h under light was determined to be approximately 1280-nt long (Figure 8A, lane a), whereas the SDGYG transcripts from the primmorphs that had been subsequently exposed for 3 h to the dark were determined to be about 1180-nt long (Figure 8A, lane b). After exposure of the primmorphs to the dark for longer than 8 h, no clear band could be detected any more under the conditions tested.
Northern blot analysis of glycogenin mRNA isolated from primmorphs that had been incubated for the indicated periods of time (in h) either in the light or dark
The length of the poly(A) tail of the SDGYG mRNA was further estimated by digesting this segment with RNase H after a pre-hybridization step with oligo(dT). Using this approach, it could be visualized by Northern blot analysis that the transcripts, which were obtained from the primmorphs kept in the light and not hybridized with oligo(dT), had an apparent length of 1280 nt (Figure 8B, lanes a and b), whereas those transcripts which had been hybridized with oligo(dT) and subsequently exposed to RNase H, in order to remove selectively their poly(A) tail, had a size of approximately 1120 nt (Figure 8B, lanes c and d).
In the last decade the molecular basis of the core oscillatory mechanisms had been elaborated, as outlined in the Introduction section and a previous review . In animals, the major ‘clock genes’ have been determined and found to code for the transcription factors BMAL1 and CLOCK; these gene products have the characteristics to act as regulatory proteins that negatively regulate their own expressions after a distinct time delay. This limb comprises a series of clock-controlled genes/gene products, e.g. Per, Cry, Ror (retinoic acid related orphan receptor)-α and nocturnin. The Rev-Erb-α gene product reduces the expression of Bmal1 via interaction with the ROR-response elements. This system is depicted in Figure 9. The phase precision of the oscillating circadian rhythm is under the control of variations in light intensities , and is likely to be adjusted by those. Surely the sponges do not have light perception systems, known from the opsin-based visual systems of protostomians or deuterostomians .
Scheme presenting the major molecules contributing to the core molecular clock machinery
It is well established that sponges undergo rhythmic contraction resulting in a reduction of their body volume by up to 70%, as shown for the species Tethya wilhelma . Interesting to note is that the contraction phases are significantly longer during the night than during the day, indicating that the sponges are provided with a nervous-like signal transmission system (e.g. ). However, in these animals, neurons are absent indicating that a nervous system, known from higher animal phyla, does not exist (e.g. ) even though neuronal receptors had been identified, e.g. the metabotropic glutamate/GABA (γ-aminobutyric acid)-like receptor  and also neurotransmitters, e.g. GABA , had been found. As a potential compensation for the lack of a neuron-based transmission system in sponges, we proposed (see the Introduction section) a neuron-like wiring circuit composed of a light-generating luciferase, that is coupled to the glass fibre network of the spicules and finally of a photon-detecting Cry.
Although the clock-controlled genes act on the expression of the central core clock molecules Bmal1 and Clock either negatively (Per and Cry) or positively (Rorα), nocturnin remains in the cytoplasm and controls the half-life and stability of key mRNA species involved in the adjustment of the homoeostasis of the intermediary metabolism, especially of the fatty acid and glycogen metabolism  (Figure 9). Nocturnin is a circadian deadenylase that had originally been described as a 3′-exonuclease from the calf thymus [17,18]. As a function for this enzyme, a selective hydrolysis of the poly(A) tail of mRNAs had been proposed , allowing a control of the post-transcriptional net polyadenylation of hnRNA (heterogeneous nuclear RNA), especially during the DNA synthesis phase. These data had been confirmed and extended for the hnRNA of nocturnin by nuclear run-on analyses . For the present study, the steady-level of the transcripts for glycogenin had been used. This enzyme is the crucial regulator of glycogen synthesis through acting both as a glycosyltransferase and as a primer .
To define the model, the response of S. domuncula and primmorphs derived from this species via aggregation of single cells to aggregates containing proliferating and differentiating three-dimensional clumps  was determined as a function of the dark and light. It is shown that primmorphs exposed to light develop into compact three-dimensional aggregates, whereas those that remained in the dark form irregular and small aggregates. If the primmorphs were kept in the dark only a slight increase in the protein and DNA contents were measured, whereas in the presence of light those aggregates vigorously form those biopolymers. Next, and as expected, the light-exposed primmorphs heavily deposited glycogen in contrast with the dark-maintained aggregates. In sponges, glycogen is produced by special cells, the grey cells , which are localized below the epithelial pinacoderm layer.
After the two cDNAs for nocturnin and glycogenin had been cloned, the deduced proteins were characterized. The nocturnin polypeptide shows the characteristic exonuclease-endonuclease-phosphatase domain , whereas glycogenin comprises the glycosyl transferase family 8 domain [45,46]. Subsequently, the nocturnin protein was recombinantly expressed in E. coli cells and then used to raise polyclonal antibodies. With these tools in hand, the expression/synthesis level of nocturnin was determined. By application of whole-mount in situ hybridization (cDNA; results not shown) and immunohistology (polyclonal antibodies; see above), the highest level of nocturnin was identified in the cortex region of the sponge. In this region, the expression of luciferase  and Cry , the two elements of the light generation and harvesting system in sponge, had also been identified. Nocturnin is abundantly present in the aquiferous canal-lining cells of the dark-kept animals, the endo-pinacocytes, as identified by immunohistology. Those cell layers had been implicated to act as a water diffusion barrier and also as phagocyting cells.
The quantitative steady-state expression of both nocturnin and glycogenin was determined by qRT–PCR, as a function of light/dark exposure. It was seen that after changing the primmorphs from the light-exposed state to the dark, a strong up-regulation of the nocturnin expression results; half-maximal levels are found after 3–6 h onset of the dark. A similar induction kinetics had been determined also for nocturnin in the mammalian system . Glycogenin gene expression was described to be induced faster after starvation for 30–60 min . In the sponge system, this enzyme reaches its 50% maximal expression level after about 6–12 h of light exposure.
In order to verify that the sponge nocturnin also posesses a poly(A)-specific 3′-exoribonuclease activity, the cDNA was used to prepare a complete nocturnin in E. coli cells recombinantly. As described for the vertebrate nocturnins [15,30], SDNOC also possesses a specific exoribonuclease activity. In our approach, the poly(A)+ glycogenin mRNA was incubated with the recombinant sponge nocturnin and the native poly(A)-specific 3′-exoribonuclease, purified from the calf thymus as described previously . Along with the data gathered from vertebrate nocturnins (summarized in ) and the results of the present study, it must be concluded that the previously described poly(A)-specific 3′-exoribonuclease, discovered by us  is identical with the nocturnin proteins.
Furthermore, based on previous studies from our group and others [48,49], it was established that the length of the poly(A) tail of mRNAs is positively correlated with the half-life of the respective transcripts . Therefore we determined the overall size of the transcript encoding glycogenin isolated from light-exposed primmorphs in comparison with the size of the transcripts from primmorphs kept, after light exposure, in the dark. It was found that the transcripts from the primmorphs under light exposure were about 100-nt longer than those isolated from the primmorphs kept in the dark. A distinct transcript band reflecting glycogenin could be identified after 3 h in the dark, whereas a longer incubation period resulted in a drastic decay of the glycogenin-specific transcripts. The conclusion that the reduction of the transcript size is due to a removal of the poly(A) tail of the glycogenin mRNA was supported by RNase H-digestion experiments, following a previously described protocol .
These digestion experiments revealed that the size of the transcripts from light-exposed animals is reduced by about 160 nt, after hybridization with oligo(dT) and RNase H treatment. The two sets of experiments strongly suggest that also in the intact cell system the size of the transcripts for glycogenin undergo a shortening of about 100–160 nt, which should be attributed to the removal of the poly(A) tail of the glycogenin transcripts. In future experiments we will use qRT–PCR under different pulse-chase conditions to determine the half-life of glycogenin mRNA isolated from the primmorphs kept in the dark in the absence and in the presence of cycloheximide, following a strategy described previously . We expect from those experiments a clarification about the effect of the inhibition of protein (nocturnin) synthesis on the rapid decay of the glycogenin transcripts. Those experiments are now in progress in our group. However, until now it has not been possible to knockout the nocturnin gene in vivo in the sponge system as described for mice , since this technique could not be established in these animals.
To conclude, the experimental results obtained in the present study show that in sponges the expression of nocturnin is up-regulated during the dark phase and undergoes decay during light exposure. The expression level of glycogenin is inversely correlated with nocturnin. On the basis of the existing data, it is reasonable to accept that the exonuclease nocturnin is, perhaps directly, involved in the dark-induced decay of the glycogenin transcripts via the deadenylation of mRNAs. The nocturnin-based and light-controlled expression system of glycogenin is economically meaningful. The sponges, as filter-feeders, filtrate huge amounts of water, 0.002–0.84 cm3 per second per cubic centimetre of sponge tissue, through their aquiferous canal system to extract edible micro-organisms and, by that, meet their extremely high energy demands. It is interesting that sponges are using the ‘high energy’ phosphate, the arginine-phosphate instead of creatine-phosphate used in mammalian systems, as the phosphoryl group donor to synthesize ATP via an enzymatic reaction catalysed by arginine kinase . In this context, the first molecular approach is outlined to understand the dark/light response at metabolic level.
brain and muscle ARNT (aryl hydrocarbon receptor nuclear translocator)-like 1
circadian locomotor output cycles kaput
heterogeneous nuclear RNA
open reading frame
quantitative real-time reverse transcription PCR
retinoic acid-related orphan receptor
Suberites domuncula glycerol 3-phosphate dehydrogenase
S. domuncula glycogenin
S. domuncula nocturnin
Werner Müller, Heinz Schröder and Xiaohong Wang designed the experiments; Vlad Grebenjuk, Michael Korzhev and Ute Schloβmacher performed the experiments; and Werner Müller, Heinz Schröder, Xiaohong Wang and Matthias Wiens wrote and/or revised the paper.
This work was supported by the ERC (European Research Council) [ERC Advanced Investigator grant number 268476 (to W.E.G.M.)], the German Bundesministerium für Bildung und Forschung (Center of Excellence BIOTECmarin project), the Deutsche Forschungsgemeinschaft [grant number SCHR 277/10-2], the European Commission [grant numbers 215507 and 286059], the International Human Frontier Science Program, the Public Welfare Project of Ministry of Land and Resources of the People's Republic of China [grant number 201011005-06] and the International S&T Cooperation Program of China [grant number 2008DFA00980].
The nucleotide sequence data reported will appear in GenBank®, EMBL, DDBJ and GSDB Nucleotide Sequence Databases under the accession number HE663082 and HE663083 for nocturnin and for glycogenin respectively.