Avarol, a cytostatically active compound from the marine sponge dysidea avara

Camp. ffiochem. Physiol. Vol. 8OC, No. I, pp. 47-52, 1985 0306~4492/85 $3.00 + 0.00 Printed in Great Britain :c: 1985 Pergamon Press Ltd AVAROL, A CYTOSTATICALLY ACTIVE COMPOUND FROM THE MARINE SPONGE DYSIDEA A VARA W. E. G. MUELLER,* R. K. ZAHN,* M. J. GA&,? N. DOGOVI~,P A. MAIDHOF,* C. BECKER,~ B. DIEHL-SEIFERT* and E. EICH~ *Institut fiir Physiologische Chemie, UniversitPt, Duesbergweg, 6500 Mainz, FRG. Telephone: (06131) 39-5911, tFaculty of Science, Department of Chemistry, Belgrade, and Institute for Marine Biology and Oceanography, Kotor, Yugoslavia and JFachbereich Pharmazie, Universitlt, Staudingerweg 3, 6500 Mainz, FRG (Received I June 1984) Abstract-l. A main metabolic product of the sponge Dysidea auara was isolated and purified and subsequently identified as avarol by applying a series of analytical techniques, e.g. [13C]NMR, [‘H]NMR and i.r. spectroscopy. 2. This sesquiterpenoid hydroquinone was found to possess strong cytostatic activity. Using the L5178y mouse lymphoma cell system in vitro (roller tube assays) avarol reduced cell growth to SOY/, at a concentration of 0.9 PM. Avarol treated cells did not show “unbalanced growth”. 3. Avarol interfered with mitotic processes, preventing telophase formation. 4. Incorporation studies with precursors for DNA, RNA, protein and glycoprotein syntheses revealed increased incorporation rates in response to avarol treatment. 5. From these results and further autoradiographicai experiments it is suggested that inhibition of cell growth is due to changes of the intracellular pools and/or alterations of the permeability properties of the cell membrane for the precursors. 6. Avarol diacetate caused the same cytostatic effect as avarol. INTRODUCTION Sponges are the simplest and oldest multicellular organisms and possess only a low grade of or- ganization. Their cells are only in very rare cases connected by junctions (Ledger, 1975); they interact mainly via soluble aggregation factors (Miiller, 1982). As in all other multicellular organisms, the establish- ment and the maintenance of the “fluid” form of a sponge individuum requires the existence of self/not-self recognition systems. One such system is based on the interaction between the species-specific aggregation factor and the inhibitory aggregation factor (Miiller et al., 1981a) and allows the recog- nition or rejection of allogeneic or xenogeneic sponge cells. Since the discovery that sponges live in sym- biosis with algae (Trigt, 1912), fungi (Roth et al., 1962), bryozoae (Wierzejski, 1935) and also with bacteria (Bertrand and Vacelet, 1971) an intense search for the biochemical basis of these relationships began. Theoretically, the symbiotic relationship be- tween sponge and non-sponge cells could be based on a positive (growth promoting) and/or on a negative principle (growth-inhibiting). In an earlier study (Miiller et al., 1981b) we could ascribe to a lectin from the sponge Halichondria panicea a growth- promoting effect for a Pseudomonas strain, isolated from the same species. The existence of growth- inhibiting activities in sponges to microbes is well proven (e.g. Burkholder and Riitzler, 1969). Address for correspondence: Professor Dr W. E. G. Miiller, Institut fiir Physiologische Chemie, Universitlt, Duesbergweg, 6500 Mainz, FRG. Analysing different sponge species for their abso- lute content of bacterial mass, we noticed that the “fluffy” sponge species, Dysidea avara, is character- ized by a comparatively low amount of bacteria (unpublished). This sponge was found to contain a sesquiterpenoid hydroquinone avarol in large quan- tities and the corresponding oxidation product ava- rone (Minale et al., 1974). However, in a subsequent contribution, these compounds appeared to be cyto- statically unimportant (Cariello et al., 1980); only at concentrations of 130 PM could an anti-proliferation effect be established in the sea urchin system. Now we show that avarol exhibits a strong cytostatic effect on growth of L5178y mouse lymphoma cells. First results on the mode of action of this compound are given. MATERIALS AND METHODS CompoundT The following labelled materials were obtained: [methyl- )H]thymidine (specific activity 14 Ci/mmole for the incorpo- ration studies and 2 Ci/mmole for the autoradiographic studies), [)H]uridine (generally labelled; sp. act. 8.4 Ci/m- mole), [3H]phenylalanine (sp. act. 9.3 Ci/mmole) and [2-‘Hlmannose (sp. act. 3.4Ci/mmole) from The Radio- chemical Centre, Amersham (UK). Isolation and pur$cation qf avurol The demosponge Dysiden avara (Schmidt) was dredged in the Bay of Kotor (Yugoslavia) in summer. The fresh material (3 kg) was homogenized in a Warring Blender in the presence of 250 ml of ethanol. Subsequently the slurry was extracted with ethyl acetate (3 x 500 ml) at room tem- perature; dry sponge residue amounted to approximately 600 g. The combined ethyl acetate extracts were washed with 47 48 W. E. G. MCJLLER et al. water (3 x 500 ml), dried over MgrSO,, filtered and evapo- described (Sachs, 1974). The size distribution curves are rated to dryness yielding 4.3 g of dark tar-like residue. This described by the following parameters as reported earlier material was taken up in benzene and applied to a 200 g (Miiller et al., 1983); mean value: the mean distribution. silica gel column (Kieselgel 60; Merck, 70-230 mesh), skewness: deviation of the distribution from symmetrical previously equilibrated with benzene. Washing with 500 ml normal (=O, negative: deviation to the right; positive: of benzene removed all non-polar components. Elution with deviation to the left), kurtosis: extent of curvature of the benzene/IO% ethyl acetate affords about 15 g of dark oily distribution with respect to normal ( = 0; negative: broader material which, according to thin-layer chromatography, than normal; positive: smaller than normal). After cdli- contains approximately 50% of avarol. Several crys- bration with mulberry pollen (diameter 12-13 pm), the tallizations from acetone-methylene chloride afford approx- mean values were converted from channel numbers to the imately 2 g of avarol. absolute cell volumes (given in pm’). The mother liquor was evaporated in uacuo, dissolved in 25ml benzene and subjected in five equal portions to high-pressure liquid chromatographic preparative purifi- cation on a Waters-500 preparative HPLC instrument using a home-made steel column containing 50g of silica gel (Kieselgel, 60; Merck, 23&400 mesh). Elution with benzene/5% ethyl acetate yielded slightly coloured material which upon crystalli~dtion from aceton~methylene chloride afforded additional 6 g of avarol; thus a total yield of avarol of 8 g (about 17% calculated on total ethyl acetate extract) was achieved. Incorporafion of nucleic acid and protein precursors For the determination of DNA, RNA, protein and gly- coprotein synthesis, 5-ml suspensions of exponentially growing cells at 100.000 cells/ml were treated for 24 hr with avarol. The labelled precursors (IO PCi each/5-ml culture) were added 24 hr or I hr prior harvest of the cultures. Samples of 1 ml were analysed for cell concentration and for acid-insoiubIe radi~dclivity (Munro and Fleck, 1966). Autoradiography Synthesis of avarol diacetare The diacetate derivative of avarol was prepared in pyri- dine with a three-fold molar excess of acetic anhydride. The product obtained was chromatographically pure and had the same characteristics as the compound described earlier (de Rosa et al., 1976). Avarone was identified in the crude extract only in minute amounts ( < 0.3”/, calculated on total ethyl acetate extract). A sample for comparison was obtained by oxidation with sitver oxide but in our hands, upon extensive purilication by column chromatography, crystalline material, m.p. 62264’C (hexane) was obtained, otherwise identical in all respect with the compound earlier described (Minale et (II., 1974). Anal_+kul me~hads Melting point was determined by using a Mettler FP52/FP5 and is uncorrected. Thin-layer chromatography was performed on precoated Merck plates of silica gel 60 F2.54 of 0.25mm layer thickness. [%]Nuclear magnetic resonance (NMR) spectra were recorded on a Joel JNM- FX60 and a Varian XL-100 spectrometer. Samples for NMR were dissolved in CDCI,. [‘H]NMR spectra were recorded in solutions (CDCI,) at 60 MHz using a Varian HA-100 spectrometer; the internal standard was tetra- methylsilane. The i.r. spectrum (KBr pellets) was measured on a Beckman spectrophotometer IR 4220 and the mass spectrum on a Varian MAT CH7A inst~ment. Ceil cullure L5178y mouse lymphoma cells were maintained in sus- pension cultures as described previously (Miiller and Zahn, 1979). For the dose-response experiments the cultures (5 ml) were initiated by inoculation of 4 x IO3 cells/ml and incu- bated at 37°C in rolier tubes for 72 hr; the controls reached cell concentrations of 5.1 x 105/ml. For some experiments 8 x IO4 cells/ml were seeded and incubation was terminated already after 24 hr; under these conditions the controls reached a density of 3.5 x 10s cells/ml. The test compounds were dissolved in O.lq6 Me,SO (final concentration); this solvent had no influence on cell growth. Cell concentration and volume distributions were deter- mined with a Model B Coulter Counter with a 32-channel size-distribution plotter (Miiller et al., 1983). The mitotic coefficient was determined after staining the cells with Harris haematoxylin and it is stated as a per- centage value (Paul, 1970). Statistical methods The ED,, was estimated by logit regression and the slope of the dose-response curve at the ED,, was calculated as Cells treated for 24 hr with the ED,, concentration of avarol (0.85 PM) were incubated for I hr in the presence of 5 pCijm1 of [3HJ~ymidine (final concentration: Z.S PM). Subsequently the cells were harvested, washed, transferred to a microscopic slide, fixed in ethanol, essentially as described (Perry, 1969). The samples were covered with a Kodak ARIO film and exposed for 5 days. Then the films were developed in a Kodak DIP developer and subsequently fixed in Agfa Acidofix. The samples were examined at unstained conditions, mounting a Schott BG38 filter be- tween light source (low voltage; 30 watts) and slide. RESULTS Since it is known from previous studies (Cariello et al., 1980; Cimino et al., 1982) that the sponge Dysidea avara contains besides avarol other structurally re- lated sesquiterpenoids and because certain symbiotic algae are also capable of producing such compounds (Fenical et al., 1973; de Rosa ef ai., 1976), we had at first to establish the purity and unequivocalIy identify the biologically active compound isolated from this sponge as avarol. Identification qf avarol Thin layer chromatographical analyses on silica gel of avarol revealed a single spot at R,= 0.00 with petroleum ether and at 0.57 with ethanol-chloroform (20:80). The spots were visualized either by 10% sulphuric acid (in ethanol) (heating at 15O’C for 15min) or by 0.5% Fast Blue Salt B (in water) (developed in ammonia at 80°C). Under U.V. light, the fluorescence quenching areas corresponded to the dye spots. Our avarol preparation had an m.p. 146148°C (not corrected); the published value is 148-150°C (CHC&) (Minale et af., 1974). The optical rotation [orlr, was also found to be +6.1”. The elemental analysis of the compound revealed the following data: determined C: 80.32 (calculated: 80.20), H: 9.81 (9.62). Our mass spectra measurement revealed: m/e 314.21 (M+, 45.5x), 191 (14.2), 124(35.9), 123 (11.0) and 95 (100). The i.r. spectrum of avarol (never published be- fore) is shown in Fig. 1. It shows no peak at a wavenumber around 1690cm-’ excluding the pres- ence of carbonylic groups. Moreover, a broad peak at 302CL3450 became visible, indicating the existence Avarol, a cytostatically active compound Wavelength pm 25 3 35 L 5 6 7 9 9 10 11 12 13 11 16 18 20 25 30 I I I t I * , I I , ,,,,,l,lil I 49 I I I I I1 I I, I I1 I I1 1 I I I I I I I I I I, LOO0 3000 2000 1500 1000 500 300 Wovenumber cm-’ Fig. 1. Infrared absorption spectrum of avarol (KBr). of phenoIic groups. These data are compatible with the published structure of avarol (Minale et al., 1974) and excludes the presence of any of the compounds described earlier from this species (Cimino et al., 1982). The sesquiterpene moiety of avarol was homolo- gized with that of avarol dimethyl ether on the basis of [13C]NMR chemical shift experiments. The chem- ical shift data of the dimethyl ether derivative were published by de Rosa et al. (1976) and are listed in brackets; C-l: 19.0 (19.5), C-2: 26.4 (26.4), C-3: 120.5 (120.5), C-4: 144.3 (144.5), C-5: 41.8 (41.8), C-6: 37.7 (37.0), C-7: 27.7 (27.6), C-8: 36.1 (36.0), C-9: 38.4 (38.2), C-10: 45.9 (45.8) and C-11: 35.8 (35.7). The signals for the oxygenated carbon atoms C-2’ and C-5’ appear at 148.7ppm, C-l’ at 126.7, while the remaining aromatic carbons are at 119.7, 116.2 and 113.9, respectively (doublets in off-resonance decou- pled spectra). The methyl group carbons (quartets) at C-4, C-5, C-8 and C-9 are visible at 19.0 (19.8), 18.1 (17.8), 17.7 (17.5) and 17.5 (17.1). The (‘HJNMR spectrum of our sample had the following characteristic signals: 6 6.58 3H, S, aro- matic; 5.16 lH (C-3H) olefinic, m; 1.02 and 1.00 overlapping 6H (8-CH, and 9-CH,). Based on these data a contamination by zonarol or isozonarol (Fenical et ai., 1973) can be excluded because of the presence of methyl groups at C-9 and C-8, as well as of the existence of the double bond between C-3 and C-4. Chromazonarol (de Rosa et al., 1976) can be excillded on the basis of the presence of the methyl group at C-8, of the oiefinic group and of a second phenolic hydroxyl group in our compound. As men- tioned, contamination of our sample with avarone and/or its derivatives is kxcluded on the basis of i.r. and U.V. spectral information. It should be mentioned that samples of potential contaminants already re- ferred to were available and their characteristics also determined. Cytostatic activity Under the experimental conditions used (roller tube technique), the natural product avarol strongly reduced cell growth (Table 1). At concentrations of 0.93 PM (72 hr incubation period) or 0.83 PM (24 hr incubation period) cell proliferation was inhibited by 50%. The activity concentration range is narrow. The following dose response effects (incubation condi- tions: 72 hr) were determined; 4 x ED,: 84% ceil growth, 2 x ED,,: 18% cell growth and 3 x ED,,: 7% cell growth. The chemically modified avarol, avarol diacetate, acted cytostatically at concentrations iden- tical to those determined for avarol itself (Table 1). The slopes of the dose response curves were calcu- lated to be parallel (significance P = 0.05). Avarol does not change the mean size of the cells signi- ficantly; the mean value of the controls was 14.50 k 5.98 (corresponding to 1183.4 + 487.7 pm’), of avarol treated cells 15.54 f 6.21 (1267.2 f 506.7) and of avarol diacetate treated cells 14.75 + 5.91 pm3 (1203.8 & 482.3). The ceil volume distribu~on curves both of untreated and of treated cultures (at E& concentration) showed a deviation to the left (posi- tive values for the skewness). Compared to the con- trols, the curvatures of the distribution curves of avarol treated celb were broader (kurtosis: -0,39), while those of avaroI diacetate treated cells were determined to be narrower (kurtosis: +0.07). Table 1. Cvtostatic activitv of avarol and in diacetate derivative ED, cont. {PM) 4000 80,OQO Parameters at the ED, cone. Compound (cells/ml) (cells/ml) Slope Mean value Skewness Kurtosis -- --~ ~--- Control - - 14.50 f 5.98 +0.64 -0.07 Avarol 0.93kO.13 0.83 + 0.1 I 1.625 15.54 + 6.21 f0.46 -0.39 Avarol diacetate 0.89+0.11 n.d. 1.414 14.75 * 5.91 +0.72 f0.07 The ED, concentrations were determined under the following two conditions; firstly, inoculation ~O~lls~rnl and incubation period 72 hr and secondly, inoculation 80,000 cells/ml and incubation period 24 hr. The different parameters, determined at the IS&,, concentration under incubation condition I, are given. n.d.: Not determined. 50 31 1 t t i 1 1 I I I 0 10 20 30 LO 50 Incubation time (hr) Fig. 2. Reversibility of the inhibition of cell growth caused by avarol. Roller tube cultures of 5 ml were imtiated by an inoculum of 8 x IO“ cells~ml. The cultures were supple- mented with 0 pM of avarol (e--O), 0.83 PM ( = ED,,; O---O), 1.66 pM ( = 2 x EL),; m-.-m) or 2.49 PM of avarol ( = 3 x ED,; x ..’ x ) and incubated for 24 hr. Then the compound was washed out; the assays were brought to a cell concentration of 8 x lO’cells/ml and incubation was continued for additional 24 hr. The mitotic coefficients of both untreated and avarol treated cells (ED,, concentration for 24 hr) were identical (6 +_ 2; N = 320). However a clear difference between the two cultures existed with re- spect to the number of dividing cells passing the telophase. While in the controls approximately 25% of the dividing cells were in the telophase, no telo- phasic cells could be detected in avarol treated cul- tures. Avarol acts cytostatically in a limited concen- tration range. As shown in Fig. 2, the inhibition of the cell proliferation with the ED,, concentration for a period of 24 hr is perfectly reversible. At concen- trations higher than 2 x EDso (3 x ED,,) the com- pound is cytotoxic for L5178y cells; after an incu- bation period of 24 hr the cell density is even 22.5% (53.7%) lower than that adjusted at the beginning of the experiment (Fig. 2). Infiuence on synthesis of macromolecules in vivo Exponentially growing cells were incubated for 24 hr in the presence of the ED, concentration of avarot. The radioactively labelled precursors for W. E. G. MCJLLER et ul. DNA (thymidine), RNA (uridine). protein (phenyl- alanine) and glycoprotein synthesis (mannose) were added 24 hr or i hr prior to harvest of the cuhures. Then the incorporation rates (acid-insoluble radio- activity) were determined. The results, summarized in Table 2, show that after incubation of the cells with avarol, a marked increase of the incorporation rate resulted. The extent of this increase was even more pronounced if the precursors were added only during the fast 1 hr of the incubation period. Autoradiographicul studies The surprising finding that L5178y ceils showed increased incorporation rates after avarol treatment required additional studies. Applying the technique of autoradio~aphy and using ~‘H]thyrnidi~~e as pre- cursor we counted 18.2 F 10.1 grainscell (N = 41) in the untreated cultures and 25.3 2 13.9 grains:‘cell (N = 35) in the avarol treated cultures. under the experimental conditions described in Fig. 3. This difference is significant at the level P = 0.025. It is conspicuous (Fig. 3) that the labelling index (ratio between iabelled and un~abe~[ed cells) is higher in control ceils (1.24) than in avarol treated cells { 1.04). DECUSSION It appears that antibiosis plays a major role firstly in maintaining balance in the marine en~~iron~nent (Kaul, 1979) and secondly has a crucial function in the establishment and also in the maintenance of symbiotic relationships between marine organisms (Miiller et al., 1981b, 1985). In the present contribution we identified that a main metabolic product of the sponge Dysidw LI~WI shows strong cytostatic activity in the mouse I.5 178~. lymphoma cell system Cr vitro (roller tube assay). The concentration of this compound in the sponge was determined to be 0.27’,!< wet wt (= I .7’:0 dry wt). By applying the techniques of [“CINMR, [‘HINMR and i.r. spectroscopy for the analysis of underivatized avarol, we could identify our compound as avarol (Minale et al., 1974). With that, we could rule out that other naturally occurring sesquiterpenoid hydro- quinones, e.g. avarone (Minale et al., 1974), zonarol and isozonarol (Fenical et al., 1973) as well as chromazonarol (de Rosa er ul., 1976) are present either as contaminant or as main component in the material we were working with. While the hitherto known sesquiterpenoid hydro- quinones seem to possess only antibacterial and/or Table 2. InRuence of war01 on the synthesis of macromolecules in exponentially growing cells Incorporation into macromoleculrs:l00.OOO cells Compound cow. [~H]Thymidine [3H]Uridine 13H]Phenyia~dnine ~~Hl.~~nno~~ (Wf (dpm) 1%) (dpm) (9,;) (dpm) i”,) (dpm) f”‘,) 0 102,100 100 27,400 100 2 1.900 I& 15.100 100 (31,300) (100) (8700) (W (6400) (100) (4200) (ItlO) 0.85 117,400 115 49,300 180 54,800 250 37.300 247 (38,800) (124) (24,400) (280) (21,100) (330) (I 1,800) (7x1) The experiments were performed as described under Materials and Methods. 0.85 PM of avaroi was added to the cultures. a concentration which reduced cell growth by 50%. The values represent means of five parallel experiments. The SD does not exceed 10%. The precursors were added either 24 hr or I hr prior IO harvcstm~ the cells; the latter values are given in parentheses. Avarol, a cytostatically active compound 51 Deducing from the chemical structure, we at first suggested that the active site of avarol is its hydro- quinone moiety which interferes with cell metabo- lism, similar to aurantiogliocladin (Vischer, 1953), by the radical intermediate semiquinone. Such an inter- mediate could be formed during an oxidation of the hydroquinone to the quinone. However, after chem- ical derivatization of avarol to avarol diacetate the strong cytostatic activity is still preserved, suggesting that it could be the sesquiterpene part of the mole- cule, rather than the hydroquinone side group of avarol, which caused inhibition of cell growth. How- ever future studies must show whether avarol di- acetate has to be enzymatically hydrolysed (in the serum or in the cell) before it can act cytostatically. Acknowledgements-This work was supported in part by a grant from the Bundesministerium fiir Forschung und Technologie (German-Yugoslav cooperation programme). REFERENCES Fig. 3. Autoradiographs of L5178y cells treated for 24 hr with 0 (A) or 0.85 PM of avarol (B). The cells were incubated with [‘Hlthymidine during the last 1 hr prior to harvesting. Subsequently the cells were processed for de- tection of incorporated [-‘H]thymidine as described under Bergman W. and Burke D. C. (1955) Contribution to the study of marine products. XXXIX. The nucleosides of sponges. J. org. Chem. 20, 1501-I 507. Bergman W. and Feeney R. J. (1951) Contribution to the study of marine products. XXXII. The nucleosides of sponges. J. org. Chem. 16, 981-987. Bertrand J. and Vacelet J. (1971) L’association entre kponges corn&es et bactkries. C.r. Acad. Sci. SK. D 273, 638-64 1. Burkholder P. R. and Riitzler K. (1969) Antimicrobial activity of some marine sponges. Nature, Lond. 222, 983-984. Materials and Methods. Magnification: x 460. anti fungal activity (reviewed by Moore, 1981), ava- rol is a first member of this class of compounds with pronounced cytostatic activity. An EDso concen- tration of 0.9 PM was determined in the L5178y cell system for avarol. Hence its activity is much higher than that determined for the classical cytostatic agents isolated from sponges, I-P-D-arabinofuranosyl- thymine and I-fl-D-arabinofuranosyluracil (Bergman and Burke, 1955; Bergman and Feeney, 1951); ED,, for ara-T: 9.8 PM (Miiller et al., 1978) and EDs,, for ara-U: 17.2pM (Miller and Zahn, 1979) both ob- tained with the same cell line. Cariello L., Giudici M. de N. and Zanetti L. (1980) Devel- opmental abberations in sea-urchin eggs induced by avarol and two congeners, the main sesquiterpenoid hydroquinones from the marine sponge @sidea aura. Comp. Biochem. Physiol. 65C, 37-41. Cimino G., De Rosa S., De Stefano S., Cariello L. and Zanetti L. (1982) Structure of two biologically active sesquiterpenoid amino-quinones from the marine sponge Dysidea auara. Experientia 38, 896. Fenical W., Sims J. J., Squatrito D., Wing R. M. and Radlick P. (1973) Zonarol and isozonarol, fungitoxic hydroquinones from the brown seaweed Dictyopreris zonarioides. J. org. Chem. 38, 2383-2386. Hubbard S. C. and Ivatt R. J. (1981) Synthesis and pro- cessing of asparagine-linked oligosaccharides. A. Rec. Biochem. 50, 555-583. Approaching the elucidation of the mode of action of avarol on cell metabolism, pulse experiments with radioactively labelled precursors for DNA, RNA, protein and glycoprotein synthesis (Munro and Fleck, 1966; Hubbard and Ivatt, 1981) were per- formed. The findings showing increased incorpo- ration rates of the precursors into macromolecules during avarol treatment suggest that inhibition of cell growth is due to changes of the intracellular pools and/or alterations of the permeability properties of the cell membrane for these precursors. Interesting is the observation that avarol interferes with mitotic processes, preventing telophase formation. Detailed studies to clarify this aspect are currently being undertaken. Because avarol does not cause un- balanced growth (Lambert and Studzinski, 1967) of the cells, it seems to be unlikely that the compound blocks cell growth during S-phase. Kaul P. N. (1979) The sea’s biomedical potential. Impact Sci. Sot. 29, 123-134. Lambert W. C. and Studzinski G. P. (1967) Recovery from prolonged unbalanced growth induced in HeLa cells by high concentration of thymine. Cancer Res. 27, 2364-2369. Ledger P. W. (1975) Septate junctions in the calcareous sponge Sycon ciliafum. Tiss. Ceil 7, 13-18. Minale L., Riccio R. and Sodano G. (1974) Avarol, a novel sesquiterpenoid hydroquinone with a rearranged drimane skeleton from the sponge Disidea aoara. Tetrahedron Lrlt. 38, 340 l-3404. Moore R. J. (1981) CRC Handbook q/’ Marine Science. Vol. II. CRC Press, Boca Raton. Miiller W. E. G. (1982) Cell membranes in sponges. Int. Rev. Cyrol. 77, 129-181. Miiller W. E. G. and Zahn R. K. (1979) Metabolism of I-p-D-arabinofuranosyluracil in mouse L5178y cells. Cancer Res. 39, 1102-l 107. Miiller W. E. G., Zahn R. K., Maidhof A., Beyer R. and Arendes J. (1978) Arabinosyl nucleosides. XII. Influence 52 W. E. G. MUELLER et al. Miiller W. E. G., Zahn R. K., Kurelec B., Lucu C., Miller of arabinofuranosylthymine on growth of L5178y cells. Chem.-biol. Inieract. 23, 141-150. Miiller W. E. G., Bemd A., Zahn R. K., Kurelec B., Dawes I. and Uhlenbruck G. (1981b) Lectin, a possible basis for K., Miiller I. and Uhlenbruck G. (1981a) Xenograft rejection in marine sponges. Eur. J. Biochem. 116, symbiosis between bacteria and sponges. J. Bacterial. 145, 573-579. 548-558. Miiller W. E. G., Geisert M., Zahn R. K., Maidhof A., Bachmann M. and Umezawa H. (1983) Potentiation of the cytostatic effect of bleomycin on L5178y mouse lymphoma cells by pepleomycin. Eur. J. Cancer clin. Oncol. 19, 665-670. Miiller W. E. G., Maidhof A., Zahn R. K., Conrad J., Rose T., Stefanovich P., Miiller I., Friese U. and Uhlenbruck G. (1985) Biochemical basis for the symbiotic relationship Dihmnim-Prochloron (Prochlorophyta). Biol Cell. 1; press. Munro H. N. and Fleck A. (1966) Recent developments in the measurement of nucleic acids in biological materials. Analysf 91, 78-87. Paul J. (1970) Cell and Tissue Culture. Livingstone. Edinburgh. Perry R. P. (1969) Autoradiography of cultured cells. In Fundamental Techniques in Virology (Edited by Habel K. and Salzman N. P.), pp. 242-249. Academic Press, New York. Rosa S. de , Minale L., Riccio R. and Sodano G. (1976) The absolute configuration of avarol, a rearranged sesqui- terpenoid hydroquinone from a marine sponge. J. chrm. Sot. Perkin I, II, 1408-1414. Roth F. J., Aheam D. G., Fell J. W., Meyers S. P. and Meyer S. A. (1962) Ecology and taxonomy of yeasts isolated from various marine substrates. Limnol. Ocean- ogr. 7, 178-185. Sachs L. (1974) Angewandte Statistik. Springer, Berlin. Trigt H. V. (1912) A contribution to the physiology of freshwater sponge (Spongillidae). Tijdschr. Ned. Dierkd. Ver. 17, l-220. Vischer E. B. (1953) The structure of aurantio- and rub- rogliocladin and glioroserin. J. them. Sot. 815-820. Wierzejski A. (1935) SiiBwasserspongien. Mim. Acra Polon. 9, I-242.