Injection of poly(b- L -malate) into the plasmodium of Physarum polycephalum shortens the cell cycle and increases the growth rate Michael Karl 1 , Roger Anderson 2 and Eggehard Holler 1 1 Institut fu ¨ r Biophysik und Physikalische Biochemie der Universita ¨ t Regensburg, Germany; 2 Molecular Biology and Biotechnology, University of Sheffield, UK Poly(b- L -malate) (PMLA) has been reported as an uncon- ventional, physiologically important biopolymer in plasmo- dia of myxomycetes, and has been proposed to function in the s torage and transport of nuclear proteins by mimicking the phospho(deoxy)ribose backbone of nucleic acids. It is distributed in the cytoplasm a nd especially in the nuclei of these giant, multinucleate cells. We report here f or the first time an increase in growth rate and a shortening of the cell cycle after the injection of purified PMLA. By comparing two strains of Physarum polycephalum that differed in their production levels of PMLA, it was found that growth activation and cell cycle shortening correlated with the relative increases of PMLA levels in the cytoplasm or the nuclei. Growth rates of a low PMLA producer strain (LU897 · LU898) were increased by 40–50% while those of a high producer strain (M 3 CVIII) were increased by only 0–17% in comparison with controls. In both strains, shortening of the cell cycle occurred to a similar extent (7.2– 9.5%), and t his w as associat ed with similar increases in nuclear PMLA levels. The effects showed saturation de- pendences with regard to the amount of injected PMLA. A steep rise of i ntracellular PMLA shortly a fter injection was followed by the appearance of histone H1 in the cytoplasm. The increase i n growth rate, the shortening of the cell cycle duration and the appearance of H1 in the cytoplasm suggest that PMLA competes with nucleic acids in binding to pro- teins that control translation and/or transcription. Thus, PMLA could play an important role in the coordination of molecular p athways t hat a re responsible f or the synchronous functioning of the multinucleate plasmodium. Keywords 1 : cell cycle; g rowth rate; P hysarum polycephalum; plasmodium; polymalic acid. In the absence of cytokinesis, repeated nuclear divisions give rise to giant multinucleate cells (plasmodia) in Physarum polycephalum [1], a well studied representative of the myxomycete family. One of the notable features of plasmodia is the high synchrony of events during the cell cycle. The maintenance of this synchrony over large cellular distances must require an activity that accounts for the rapid and ubiquitous distribution and coordination of protein activities in the periodical cell cycle eve nts. W e h ave previously identified the unusual polyanion poly(b- L -ma- late) (PMLA) a s a specific component of the plasmodium that fulfils the requirements for such a Ôdistributing activityÕ [2,3]. Its level in the nuclei i s kept constant by constitutive synthesis and secretion of excess polymer from the cyto- plasm to the culture medium, and the levels in the nuclei for different strains are of the same magnitude [4]. PMLA binds reversibly to histones, DNA polymerases, a nd other DNA- interacting proteins, thus favouring the formation of large complexes consisting o f a variety of proteins. The binding involves specifically the array of negative carboxylates on the PMLA chain that is isosteric with the array of phosphates in nucleic acids [2,5–9]. This complex-forming property and the high mobility seen for the fluorescently labelled polymer in plasmodia [10] suggest that PMLA could function as a constituent of the postulated distributing activity. In addition to functioning as a distributing activity, PMLA might act as a synchronizing agent by competing with nucleic acids for the binding of structural proteins, enzymes, and regulatory proteins. A recent analysis of DNA synthetic activities in extracts of plasmodia revealed a cell cycle dependent inhibition and activation of DNA poly- merases. This could be explained by the binding of DNA polymerases to endogenous PMLA in competition with periodically synthesized histones or certain other proteins [11]. Competition of this kind is likely to inhibit various kinds of activities involving the binding of proteins to nucleic acids, and it c ould affect cell g rowth a nd cell cycle duration. The distributing activity of PMLA and the efficiency of competition between PMLA and nucleic acids would both be influenced by the concentration of PMLA. An abnormal increase in PMLA level would therefore be expected to modulate growth properties. To test this prediction, we injected purified PMLA into plasmodia and measured the cytoplasmic and nuclear levels of the polymer in parallel with changes in growth rate and cell cycle duration. The Correspondence to E. Holler, Institute fu ¨ r Biophysik und Physikalische Biochemie der Universita ¨ t Regensburg, D-93040 Regensburg, Germany. Fax: +49 941943 2813, Tel.: +49 941943 3030, E-mail: eggehard.holler@biologie.uni-regensburg.de Abbreviations: PMLA, poly(b- L -malic acid) and poly(b- L -malate). Enzyme: phosphatase (EC 3.1.3.16). Note: A website is available at http://www.biologie.uni-regensburg. de/biophysik (Received 2 5 May 2004, revised 5 July 2004, accepted 23 July 2004) Eur. J. Biochem. 271, 3805–3811 (2004) Ó FEBS 2004 doi:10.1111/j.1432-1033.2004.04299.x investigation included wild-type and mutant strains of P. polycephalum with distinctly different levels of PMLA in the cytoplasm, but more or less comparable levels in the nuclei. We found that increased cytoplasmic and nuclear levels of PMLA induced a strain specific enhancement of cell growth and an equal (between strains) shortening of the cell cycle. Materials and methods Strains and materials The following strains of P. polycephalum were used (Table 1): t he Ôhigh PMLA producingÕ strains M 3 CVIII ATCC 96951 (yellow, wild-type), M 3 CVII ATCC 204388, CH813 · LU861 (yellow), LU688 (yellow); the Ômedium PMLA producingÕ strain OX 110 · RA271 (yellow); and the Ôlow PMLA producingÕ strains LU897 · LU898 (white, mutant) [ 12], and LU887 (white, mutant). Plasmodia were routinely g rown for 1.5–2 days at 24 °C, except M 3 CVIII, which was grown at 27 °C.Theaxenicgrowthmediumhas been described [13]. Macroplasmodia were started from the fusion of microplasmodia (15–20 mg) on the surface of 2% (w/v) agar in 8-cm plastic Petri-dishes [14]. PMLA, potas- sium salt, was purified as described [15], having M r ¼ 50 kDa and a polydispersity ¼ 2.0 [16]. Anti-(histone H1) Ig (bovine) raised in sheep (diluted 1 : 500 ELISA), anti- (histone H2B) ( bovine) raised in sheep (1 : 2000), anti- (histone H3) Ig (bovine, subgroup f3) raised in sheep (1 : 2000), anti-(histone H2A + H4) (calf thymus) raised in sheep (1 : 1 000) were all from B ioTrend (Cologne, Ger- many) 2 . Peroxidase-coupled anti-IgG (sheep) raised in rabbit was from Calbiochem. All anti-histone immunoglobulins bound to histones purified from P. polycephalum [17] in ELISA and Western blotting. The specificity was charac- terized by Western blotting and ELISA of purified total histones and nuclear and chromatin extracts of P. poly- cephalum. A low d egree of cross reaction of anti-H1 Ig with the core histones was observed that was constant and negligible during ELISA under t he conditions used; other- wise, the immunochemical responses of the a ntibodies were specific. Lambda protein phosphatase (400 000 unitsÆmL )1 ; > 300 000 UÆmg protein )1 ) w as from Calbiochem. A ll other reagents were f rom Merck or Sigma and were of the highest purity available. Microinjection, growth rate Before each macroplasmodium received a microinjection, a small piece was cut out and the stage of the cell c ycle determined as described previously [18]. The stages could be fairly well predicted by c alculation, taking into account the time elapsed after the fusion and the known length of the cell cycle. The injection solution of 1–4 lL contained 15–200 mgÆmL )1 PMLA, p otassium salt, or a reference solution containing either KCl, L -malate (potassium salt), poly( L -glutamate) (potassium salt) or distilled water. Solu- tions were injected into veins in parallel to the protoplasmic streaming at fi ve different points distributed over the plasmodium. Fluorescently l abelled PMLA [ 10] s howed an even distribution over > 97% of the plasmodium within less t han 20 min. A Leitz 3 micromanipulator (Wetzlar, Germany) equipped with a laboratory course binocular (Wild Heerbrugg, Heerbrugg, Switzerland) 4 and Kwik-Fil TM borosilicate capillaries (World Precision Instruments, Sarasota, FL, USA) 5 were used. Times of injections were either early S -phase following the third mitosis or early G 2 -phase following the second mitosis. Sizes of plasmodia were then of the order of 4–7 cm 2 . The third metaphases were observed at 25.6 ± 0.5 h (mean ± SD, 10 replicates) after the fusion of microplasmodia for the yellow M 3 CVII strain, at 2 5 ± 0.5 h for t he white mutant strain LU897 · LU898, and at 21 ± 0.4 h for the white mutant strain LU887. Cell growth was measured at various times and interpolated for 5 h after the injectio n. The duration of the nuclear division cycle was measured microscopically [18] between the third and the fourth mitosis, when at least 60% of the nuclei were in metaphase. To follow the growth of a p lasmodium in a noninvasive manner, its surface area w as measured at successive times [19]. The plasmodia did not contact the walls of the agar plates at any time. The s urface areas correlated significantly with the weight of (wet) plasmodia measured after their removal from the agar plates. One square centimete r corresponded to 18.3 ± 0.5 m g plasmodium (mean ± SD, six independent measurements). Table 1. PMLA contents and numbers of nuclei in various strains of P. polycephalum. All results are given in means and standard deviations of at least three independent measurements. Strain Colour (genotype) a PMLA contents (lgÆg plasmodium )1 ) in extracts of: Number of nuclei (10 8 Æg plasmodium )1 ) Nuclei Cytoplasm Culture medium M 3 CVII Yellow (whiA + /whiA + ) a 200 ± 75 b 60 ± 15 b High 2.5 ± 0.4 c 450 ± 45 c 350 ± 25 c M 3 CVIII Yellow (whiA + /whiA + ) 189 ± 20 b 63 ± 14 b High 2.0 ± 0.5 c 340 ± 28 c 350 ± 25 c CH813 · LU861 Yellow (whiA + /whiA + ) 550 ± 60 b 460 ± 90 b Very high 5 ± 1 b LU688 Yellow (whiA + ) 1730 ± 150 b 1000 ± 60 b High 4.8 ± 1 b OX110 · RA271 Yellow (whiA1/whiA + )18±3 b 47 ± 15 b Low 0.2 ± 0.05 b LU897 · LU898 White (whiA1/whiA1) 270 ± 25 b Not detectable b Not detectable 1.8 ± 0.2 b LU887 White (whiA1) 130 ± 16 b Not detectable b Not detectable 2.0 ± 0.5 c 260 ± 15 c 60 ± 15 c a Presumed genotype. b Microplamodia. c Macroplasmodia: G 2 -phase. 3806 M. Karl et al. (Eur. J. Biochem. 271) Ó FEBS 2004 Quantitative ELISA Extracts of cytoplasm, nuclei, and chromatin were diluted 10 4 )10 8 -fold in buffer (1.59 g Na 2 CO 3 , 2.93 g NaHCO 3 in 1L H 2 O, pH 9.6) and coated in varying amounts onto microwell plates for 3 h at 37 °C. The plates were washed with phosphate/saline (2 m M KH 2 PO 4 ,8m M Na 2 HPO 4 , 137 m M NaCl, 4 m M KCl pH 7.4), blocked o vernight in a solution of 2.5 mgÆmL )1 milk powder in phosphate/saline and 0.05% (v/v) Tween 20 at 4 °C, followed by f our changes of washing solution, containing 0.9% (w/v) NaCl, 0.05% (v/v) Tween 20, before incubation with 50 lLper well anti-histone Igs for 1.5 h at 37 °C. After three changes of washing buffer, the second antibody (peroxidase conju- gated) was administered a t 50 lLperwellfor1hat37°C. Following four changes of washing buffer, plates were incubated with 50 lL per well of a solution of (o-phenylene- diamine dihydrochloride) Fast Tablets TM (Sigma) for 25 m in at room temperature in the dark. A fter the addition of 16 lLperwellof3 M HCl, the extinction was read at 490 n m and reference 630 nm wavelengths i n a Micro- Reader Dynatech MR700 6 (Dynatech Laboratories Inc., Chantilly, VA, USA). The readings were plotted as a function of A 595 units (protein assay according to Bradford [20]). Increments of the linear fits through t he origin were used to calculate the relative antigen concentrations in units of 10 5 · A 490 /A 595 . The ELISA for H1 depended on the degree o f phosphorylation, and no a ttempts were made to standardize the ELISA readings on an H1 mass basis. Series of measurements were compared after standardization with a reference sample prepared by the same method. PMLA in the extracts did not affect ELISA readings. Other methods Cytoplasmic, nucleoplasmic and chromatin extracts were prepared according to a modified method of Angerer a nd Holler [6]. One gram of plasmodia was lysed in 11 mL of homogenization buffer and centrifuged for 10 m in at 2000 g. The supernatant was removed and centrifuged again at 20 000 g; the resulting supernatant w as the cytoplasmic extract. The pellet was treated with nuclei extraction buffer [6] and centrifuged at 20 000 g;the supernatant was referred to as the nuclear extract. The residual pellet was incubated with an equal volume containing sodium carbonate buffer and 10 M NaCl, pH 11.4; the supernatant after 20 000 g centrifugation was the chromatin extract. This fractionation system was optimal with regard t o the preservation of nuclei during the preparation of the cytoplasmic extract, and the same r esults were obtained with t he homogenization buffer of Loidl and Gro ¨ bner [21]. Nuclei were counted in the pellets using a Neubauer hemocytometer before extraction for calculation of the PMLA content. The purification of nuclei over a Percoll gradient was not used, as nuclei prepared in this way were devoid of PMLA [8]. For whole plasmodia, nuclei were counted in alcohol-fixed smears under the phase contrast microscope or after staining with 4¢,6¢-diamino-2-phenyl- indole. Plasmodium mass was assessed after removal of adhering liquid with tissue paper. Values are given as means ± SD for measurements p erformed at least i n triplicate. To dephosphorylate histone H1, extracts were incubated for 30 m in at 30 °C in the presence of 800 U Lambda protein phosphatase, 50 m M Tris/HCl pH 7.5, 5 m M dithiothreitol, 2 m M MnCl 2 ,and100lgÆmL )1 BSA. Poly- malate was quantitated as described by Karl et al. (2003) [10]. DNA was measured by the me thod described by Gold and Shochat (1980) [22]. Results and Discussion Growth rate and duration of cell cycle In the intial experiments, PMLA was injected into macro- plasmodia, and growth was measured in comparison to control plasmodia injected with water (Ômock-injectedÕ plasmodia). The yellow strain M 3 CVIII produces and secretes high amounts of PMLA, while the white strain LU897 · LU898 produces considerably less PMLA (Table 1). Both strains grew very slowly during the first few hours following fusion of microplasmodia, before they assumed an approximately constant or slightly exponential growth rate. E ighteen hours after fusion, each plasmodium was injected with 200 lgPMLA(Ômock injectedÕ plasmodia received an equal volume of water), and growth was allowed to continue until 45 h after fusion (Fig. 1). Sizes (corres- ponding to masses, see Materials and methods) were measured 5, 20, and 27 h after injection. The y ellow plasmodia grew faster than the white ones, but PMLA- injected plasmodia of both strains grew at comparable rates. After 45 h, the sizes of the PMLA-injected yellow plasmo- dia were larger by 20% (P<0.007) and the PMLA- injected white plasmodia were larger by 44% (P<0.004) than their water-injected control counterparts. Noninjected plasmodia grew at the same rates as the water-injected controls (data not shown). Thus, the effect of PMLA on growth was greater on the low PMLA producing white plasmodia, which c ontained l ess P MLA, than on the h igh PMLA producing yellow plasmodia (Table 1). The greater sizes persisted until the end of the experiment. We then tested whether the plasmodial size (growth rate, Fig. 2A,C,E) and the duration of the cell cycle (Fig. 2B,D,F) depended on the dose of injected PMLA, over the range 0–400 lg PMLA (A–D). The times of the injections were early G 2 -phase before mitosis III (Fig. 2C,D) or early S-phase, following mitosis III (Fig. 2A,B,E,F), in order to see whether mitosis influenced the efficacy of the PMLA- induced phenomena. Plasmodial sizes were measured at 5 h after injection (Fig. 2A,C,E) and cell cycle duration between the third a nd fourth mitosis (Fig . 2 B,D,F). The ratio of cell sizes in the figures is expressed relative to control (mock ¼ water-injected) plasmodia a nd reflects the growth rate. To test for specificity, 0–300 lg poly( L -glutamate) were injected in control experiments (Fig. 2E,F). Poly( L -glutamate) was chosen as a negative control because it is not isosteric with the (deoxy)ribosephosphate backbone of nucleic acids [5]. Control e xperiments to test for osmotic effects were carried out with KCl and L -malate (potassium salt) at equimolar amounts with the malyl residues of injected PMLA. The PMLA-injected plasmodia grew to be larger than the mock-injected ones, with a dose–response relationship indicating that the system w as approaching saturation at high doses of PMLA (Fig. 2A,C). The low PMLA producer Ó FEBS 2004 Polymalate effects growth and cell cycle (Eur. J. Biochem. 271) 3807 (mutant white strain LU897 · LU898, solid lines in Fig. 2) showed a greater increase in size than the high PMLA producer (yellow strain M 3 CVIII, broken line in Fig. 2). Thus, after injection of 400 lg PMLA in S-phase, the yellow plasmodia were larger by 17% (P<0.015) that their mock-injected contro ls (Fig. 2A, dashed line) and not detectably larger after injection in G 2 -phase (Fig. 2C, dashed line). In contrast, the white plasmodia were larger by 50% (P<0.0003) than the mock-injected controls after injection in S-phase (Fig. 2A, continuous line), and larger by 40% (P<0.0006) after injection in G 2 -phase Fig. 2C, continuous line). The differences between the strains are highly significant and are in agreement with those in Fig. 1. The cell cycle duration decrease d significantly in com- parison with the mock-injected plasmodia, the overall changes being similar for the two strains a nd indepen dent of the injection times. Following injection of 400 lgPMLA, the decrease was 8.2% (P<0.0012) f or yellow plasmodia injected in S-phase or G 2 -phase (Fig. 2 B,D, dashed lines), and 9.5% (P<0.0003) and 7.2% (P<0.0005) for white plasmodia injected in S-phase or G 2 -phase, respectively (Fig. 2 B,D, continuous lines). The saturation behaviour of the dose dependence was less pronounced than that of the plasmodium size. The effects of injecting at different time s (S- and G 2 -phase) were similar and did not give evidence for a control point, except for the failure of yellow plasmodia to increase in size after injection in early G 2 -phase. This could indicate that during the 5 h after injection in early G 2 -phase, growth stimulation was low because plasmodia a lready contained high levels of PMLA. To examine the effect on cellular p rotein and DNA concentrations, 200 lg PMLA was injected in early S-phase (after mitosis III) and the contents compared with those o f mock-injected plasmodia. After 7 h, the protein contents (mgÆg plasmodia )1 ) were: 10.9 ± 1.2 (mock-injected) and 11.1 ± 1.2 (PMLA-injected) for the yellow strain M3CVIII; and 14.6 ± 1.5 (mock) and 14.9 ± 1.7 (PMLA) for the white strain LU897 · LU898. The DNA contents (lgÆg plasmodia )1 ) were: 522 ± 18 (mock) and 540 ± 18 (PMLA) for the yellow strain; and 690 ± 12 (mock) and 744 ± 30 (PMLA) for the white s train. Thus, the concen- trations for PMLA- and mock-injected plasmodia were not significantly different (P > 0.05 at 95% confidence levels). Also injections of either L -malate, KCl or poly( L -glutamate), potassium salt (Fig. 2E,F) at relevant concentrations had no significant effect on size or ce ll cycle duration (P >0.05 at 95% co nfidence levels). These results showed that the effects were specific and not caused by an increased osmolarity or availability of malate as a metabolite. Distribution kinetics of injected PMLA A comparison o f t he results i n Figs 1 and 2 (A,C) reveals that the low PMLA-producing w hite strain manifests a larger increase in the growth rate than the high PMLA- producing yellow strain, while almost no difference between the strains is seen for the cell cycle duration. In the following discussion, these phenomena will be related to t he changes in the amounts of PMLA in cytoplasmic and nuclear extracts. As shown in Table 1, the PMLA contents of nuclei extracted f rom the high PMLA producing (M 3 CVIII) and the low PMLA producing (LU897 · LU898) strains are comparable, while the PMLA concentrations in cytoplasmic extracts differ considerably. A similar relationship is seen for strains M3CVII and LU887, which also represent high and low producers of PMLA (Table 1 ). In order to study changes in the levels of PMLA in cytoplasm and nuclei, we injected 400 lg of the polymer into M3CVII and LU887 plasmodia (weights of 150 mg) and measured the PMLA contents of nuclear and cytoplasmic extracts. The results in Fig. 3A,B for the low PMLA-producing strain LU887 show an immediate increase after injection, peaking at approxi- mately 2.8 mgÆgplasmodium )1 in the nuclear extract (Fig. 3 A) and at 350 lgÆg plasmodium )1 in the cytoplasmic extract (Fig. 3B). The PMLA contents in mock-injected or noninjected plasmodia remain constant (Fig. 3B and [3]). Similar increases were found for the yellow strain M 3 CVII (data not shown). Thus, the absolute increases in PMLA levels in the extracts were the same for the white and the yellow strains. N evertheless, because the level in the cytoplasm of w hite plasmodia before injection was v ery low (Table 1), the relative increase in these str ains was considerably higher than for the yellow strains, which Fig. 1. Effect of i njected PMLA on gro wth. Volumes of 2 lLPMLA solution (200 lg PMLA, dashed lines) or distilled water (mock-injec- ted, continuo us lines) were injected at the time indicated by the arrows. The size of p lasmodia was measured in terms of the surface area covered by each plasmodium (1 cm 2 corresponding to 18.3 ± 0.5 mg plasmodium, see Materials and methods). (A) Yellow wild-type strain M 3 CVIII. (B) White mutant strain LU897 · LU898. Standard devi- ations of three independent measurement s are indicated. 3808 M. Karl et al. (Eur. J. Biochem. 271) Ó FEBS 2004 already had high levels of cytoplasmic PMLA before injection. Assuming similarity between the two white strains on one hand and the two yellow strains on the other the different relative increases of PMLA in the cytoplasmic extracts from yellow and white plasmodia can be seen to correlate qualitatively with the different increases in growth rates (Fig. 2A,C). In contrast, the PMLA contents of nuclear extracts from yellow and white plasmodia are similar (Table 1), and this correlates with the almost eq ual degree of shortening of the cell cycle (Fig. 2B,D). The kinetics of PMLA distribution were remarkable. (a) PMLA contents increased rapidly after injection, in agree- ment with previous findings [10]. A calculation shows t hat the injected 400 lgPMLA(into 150 mg p lasmodium) was recovered in the extracts. (b) The amounts of free L -malate in the cytoplasmic/nuclear extracts were 260 ± 30 lg/45 ± 8 lg and constant over the period of the i nvestigation (data not shown). PMLA w as not detectably degraded, but was instead secreted into the culture medium [3,10]. ( c) Clearance from the plasmodium in 4 h (Fig. 3A,B) corresponded to a clearance activity of 600 lgPMLAÆh )1 Æg plasmodium )1 and compares with the rate of 920 lgÆh )1 Æg plasmodium )1 secretion by micro- plasmodia (60 h after inoculation) [23]. These results are in agreement with the homeostatic model described previously [3] and show that plasmodia do not tolerate artificially increased levels of PMLA. The data in Table 1 allow calculation and comparison of the PMLA contents of nuclei in a variety of strains. These nuclear contents vary relatively little, between 0.65 pmol for LU887 and 3.6 pmol for LU688, whereas the variation in cytoplasmic contents is much greater (> 1000-fold). This suggests that the nuclear concentration of PMLA is regulated to maintain a relatively constant, high level in all strains. Fig. 2. Effects of polymalate injection on growth and duration of the cell cycle. Variable amounts of PMLA (A–D) or poly( L -glutamate) (E,F) were injected in early S-phase (40 ± 5 min afterthethirdmetaphase;A,BandE)orinG 2 -phase (180 ± 10 min after the second metaphase; C,D,F). M 3 CVIII (yellow, high PMLA producer; j) and LU897 · LU898 (white, low PMLA producer; d). The sizes of plasmodia were measured after 5 h and the cell cycle duration between the third and fourth mitoses. The sizes are given relative to those of mock-injected plasmodia, which were grown in parallel, and indicate the ratio of growth rates (Fig. 1). S tandard deviations refer to at least three independent measurements. Ó FEBS 2004 Polymalate effects growth and cell cycle (Eur. J. Biochem. 271) 3809 Transient appearance of histone H1 in the cytoplasm We suspected that the mechanism(s) underlying the increase in the growth rate and the shortening of the cell cycle might be related to the ability o f PMLA to form c omplexes with nucleic acid binding proteins and thus to compete with nucleic acids. Known e xamples are histones, especially H1 [5,6]. Therefore we tested whether free H1 and core histones (probably in complexes with PMLA) could be d etected as a consequence of the injection o f PMLA. Using quantitative ELISA on cytoplasmic extracts (Fig. 3C), increase d levels of H1 were indeed detected. The increase followed the injection of PMLA with a delay of 1–2 h. As epitopes had been masked by phosphorylation in vivo, a higher amount of H1 was detected after dephosphorylation with Lambda protein phosphatase (Fig. 3C, j). Injection of L -malate did not provoke H1 appearance (Fig. 3, h). ELISA with specific antibodies did not detect an increase of the core histones (data not shown). Also, the nuclear extract and chromatin extract did not indicate a PMLA-dependent variation in histone content. Conclusion Our main finding was that plasmodia responded t o an artificial, unprogrammed increase in t he cellular content of PMLA by an in crease in size (growth r ate) and by a shortening of the cell cycle duration. This conclusion is based on the study of more than 120 plasmodia, which gave relatively high experimental reproducibility a nd results that are statistically highly significant. The effect was structure specific for PMLA. Swelling of plasmodia or an accumu- lation of slime after injection could be ruled ou t as explanations, b ecause concentrations of protein a nd DNA remained unaffected. Th e possibility that PMLA m ight be hydrolysed to L -malate and th en used as a carbon and energy source was e xcluded on the basis t hat not only was PMLA hydrolysis absent (this work and [3,10]) but that injection of L -malate did not reproduce the effect. It was concluded that acceleration o f the cell cycle a nd enhanced growth were functional effects of injected PMLA. We propose that the underlying mechanism by which PMLA increases the growth rate and shortens the duration of the cell cycle is the polymer-inherent isosterism of the carboxylates with the array of phosphates in nucleic acids and, consequently, the competition with nucleic acids in the (reversible) binding of p roteins, such as histones and DNA polymerases [5,6,10]. An example of competition between DNA polymerases and histones in the binding of PMLA has been demonstrated in a purified system [2] and was suggested as an explanation for the periodic activation of DNA polymerases in the cell cycle [11]. The degree of competition depends on the concentration of free PMLA, DNA, and the binding affinity and follows saturation functions. It i s speculated that the levels of protein complexes of intrinsic PMLA and nuc leic acids are ÔtunedÕ for optimal function in the plasmodium and that an unscheduled increase in PMLA by injection perturbs this tuning and causes the observed effects. Different proteins and nucleic acids gave rise to the different dose depend- encies in Fig. 2. For instance, in competing with mRNA for t he binding of regulatory proteins in the cytoplasm, PMLA might derepress translation activity. Examples of translational regulation are known for higher eukaryotes: mRNAs are masked during gametogenesis until embryonic development [24]. In another example, the expression of maternal proteins is suppressed in mouse oocytes by the binding of MSY2 protein to mRNA [25]. Also, PMLA might b ind to histones a nd enhance the rates of transcrip- tion by facilitating chromatin remodelling. Our finding of high levels of free (probably PMLA-bound) H1 in the cytoplasm after injection supports this assumption. High concentrations of (phosphorylated) linker histone H1 in the cytoplasm during S-phase and G 2 -phase of the cell cycle but not in G 1 -phase have been reported for HeLa cells [26]. In some other mammalian cells, examples of cytoplasmic accumulation of H1, but not of core histones have been reported [27]. Fig. 3. Levels of polymalate and histone H1 during the cell cycle after injection of 400 lg PMLA per macroplasmodium (white mutant strain LU887,  150 mg) in early S-phase (arrow). (A) PMLA in the nuclear extracts. (B) PMLA in the cytoplasmic extracts; j, noninjected macroplasmodia. ( C) E LISA o f h istone H 1 in cytoplasmic extracts without (m) and after incubation with Lambda protein phosphatase (j);onerelativeunit¼ 10 5 · A 490 /A 595 (A 490 ELISA readings, A 595 protein readings according to th e assay described by B radford [20], see Materials and methods). Controls refer to macroplasmodia having received 400 lg L -malate (h) and to no ninjected m acroplasmodia (s). The symbo l M3 refers to the t hird mitosis after the fusion of micro- plasmodia. Mean values and SDs of at least three i ndependen t meas- urements are indicated. 3810 M. Karl et al. (Eur. J. Biochem. 271) Ó FEBS 2004 One may wonder why the large increase in PMLA content after injection w as par alleled b y only modest changes in growth and cell cycle duration. However, we do not know how great an effect should be expected. If we assume that certain PMLA–protein complexes are involved in growth and cell cycle timing, then the size of the observed effect will depend on the amount of such newly formed PMLA–protein complexes. If the proteins are already almost completely saturated by intrinsic PMLA before the injection, even large amounts of injected PMLA will not give rise to significant increases in the levels of these complexes and thus will not induce large effects. On the basis of these considerations, the greater effect on growth in Fig. 2A,C can be explained by assuming lower levels of c ertain PMLA–protein complexes in white p lasmodia than in yellow plasmodia before the injection ( because t here is much less PMLA in the cytoplasm of white plasmodia; Table 1). By the same token, the equal (between-strains) effects on the cell cycle duration (Fig. 2B,D) might reflect almost identical levels of certain PMLA–protein complexes in the nuclei of w hite and y ellow strains (containing similar a mounts of PMLA; Table 1 ). The results provide evidence not only that PMLA functions as a storage and carrier molecule for certain proteins [6], but also that it may be involved in molecular events concerned with growth and cell cycle duration in plasmodia. Because t hese events are synchronized, PMLA may a lso p lay a role in synchronization. The f undamental mechanism of its function is mimicry of the charge array on the nucleic acid backbone and competition with nucleic acids for binding to specific proteins. A full understanding of this mechanism remains to be established. References 1. Burland, T.G., Solnica, K.L., Bailey, J., Cunningham, D.B. & Dove, W.F. (1993) Patterns of i nheritance, development and the mitotic cycle in the p rotist Physarum polycephalum. Adv. M icrob. Physiol. 35, 1–69. 2. Fischer, H., Erdmann, E.S. & Holler, E. (1989) An unusual polyanion from Physarum polycephalum that inhibits homologous DNA polymerase a in vitro. Biochemistry 28, 5219–5226. 3. Sc hmidt, A., Windisch, C. & Holler, E. (1996) Nuclear accumu- lation and homeostasis of the unusual polymer b-poly( L -malate) in plasmodia of Physarum polycephalum. Eur. J. Cell. Biol. 70 , 373–380. 4. Rathberger, K ., Reisner, H., Willibald, B.H., Molitoris, H P. & Holler, E. (1999) Comparative synthesis and hydrolytic degrada- tion of poly( L -malate) by myxomycetes a nd fu ngi. Mycol. Res. 103, 513–520. 5. Holler,E.,Achhammer,G.,Angerer,B.,Gantz,B.,Hambach,C., Reisner, H., Seidel, B., Weber, C., Windisch, C., Braud, C. & Vert, M. (1992) Sepcific inhibition of Phys arum polycephalum DNA- polymerase-a-primase by poly(b- L -malate) and related polyani- ons. Eur. J. Biochem. 206,1–6. 6. Angerer, B. & Holler, E. (1995) Large complexes of beta-poly ( L -malate) with DNA polymerase alpha, histones, and other proteins in nuclei of growing plasmod ia of Physarum poly- cephalum. Biochemistry 34 , 14741–14751. 7. Achhammer, G., W inkler, A., Angerer, B. & H oller, E. (1995) DNA polymerase d of Physarum polycephalum. Curr. Genet. 28, 534–545. 8. Doerhoefer, S ., Khodyre va, S., Safronov, I.V., W lassoff, W.A ., Anarbaev, R., Lavrik, O.I. & Holler, E. (1998) Molecular con- stituents of the replication apparatus in the plasmodium o f Physarum polycephalum: identification by photoaffinity labelling. Microbiology 144, 3181–3193. 9. Karl, M. & Holler, E. (1998) Multiple polypeptides immuno- logically related to b-poly( L -malate) h ydrolase (polymalatase) in the plasmodium of the slime mold Physarum polycephalum. Eur. J. Biochem. 251, 405–412. 10. Karl, K., Gasselmaier, B., Krieg, R. & Holler, E. (2003) Locali- zation of fluorescence-labeled poly(malic acid) to the nuclei of the plasmodium of Physarum polycephalum. Eur. J. Biochem. 270, 1536–1542. 11. Do erhoefer, S., Windisch, C., Angere r, B., Lavrik, O.I., Lee, B S. & H oller, E. (2002) The D NA-polymerase inhibiting activity of poly ( b- L -malic acid) in n uclear extr act during the cell cycle of Physarum polycephalum. Eur. J. Biochem. 269, 1253–1258. 12. Anderson, R.W. (1977) A plasmodial colour mutation in the myxomycete Physarum polycephalum. Genet. Res. 30, 301–306. 13. Daniel, F.W. & Baldwin, H.H. (1964) Methods of culture for plasmodial myxomycetes. Methods Cell Physiol. 1, 9–14. 14. Nygaard, O.P. & Guttes, S.R.H.P. (1960) Nucleic acid metabolism in a slime mold with synchronous mitosis. Biochim. Biophys. Acta 38, 298–306. 15. Holler, E. (1997) Pol y(malic acid) f rom natural sources. In Handbook of Engineering Polymeric Materials. (Cheremisinof, N.P., ed.), pp. 93–103. Marcel Dekker, New York. 16. Lee, B S., Vert, M. & Holler, E. (2002) Water-soluble aliphatic polyesters: poly(malic acid)s. I n Biopolymers 3a (Doi,Y.& Steinbu ¨ chel, A., eds), pp. 75–103. Wiley-VCH, Weinheim. 17. Albert, P., Lacorre-Arescaldino, I. & Toublan, B. (1993) High resolution polyacrylamide gel electrophoresis of Physarum po ly- cephalum hist ones: improved resolution of variants and mod ified forms. Anal. Biochem. 209, 224–227. 18. Mohberg, J. (1982) Recognition of mitosis. In Cell Biology of Physarum and Didymium (Aldrich, H.C. & Daniel, J.W., eds), pp. 273–276. Academic Press, New Y ork. 19. Wright, M ., Lacorre-Arescaldino, I., Macquet, J.P. & Daffe ´ ,M. (1984) Introduction o f polyploid nuclei in the plasmodium of Physarum polycephalum by antitumor compounds. Cancer Res. 44, 777–783. 20. Bradford, M.M. (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. B iochem. 72, 248–254. 21. Loidl, P. & Gro ¨ bner, P. (1987) Histone synthesis during the cell cycle of Physarum polycephalum. Synthesis of different histone species is not under a common regulatory control. J. Biol. Chem. 262, 10195–10199. 22. Gold, D.V. & Shochat, D. (1980) A rapid colorimetric assay for the est imation of microgram quantities of DNA. Anal. B iochem. 105, 121–125. 23. Lee, B S . & Holler, E. (1999) Effe cts o f culture cond itions on b-poly( L -malate) produc tion by Physarum polycephalum. Appl. Microbiol. Biotechnol. 51, 647–652. 24. Sommerville, J. & Ladomery, M. (1996) Transcription and masking of mRNA in germ cells: involvement of Y-box proteins. Chromo so ma 104, 469–478. 25. Yu, J., Hecht, N.B. & Schultz, R.M. (2002) Expression of MSY2 in mouse oocytes and preimplantation embryos. Biol. Reprod. 65, 1260–1270. 26. Bleher,R.&Martin,R.(1999)Nucleo-cytoplasmictranslocation of histone H1 during HeLa cell cycle. Chromosoma 108, 308–816. 27. 7 Zlatanova, J.S., Srebreva, L.N., Banchev, T.B. & Tsanev, K.G. (1990) Cytoplasmic pool of histone H1 in mammalian cells. J. Cell Sci. 96, 461–465. Ó FEBS 2004 Polymalate effects growth and cell cycle (Eur. J. Biochem. 271) 3811 . Injection of poly(b- L -malate) into the plasmodium of Physarum polycephalum shortens the cell cycle and increases the growth rate Michael. rate and shortens the duration of the cell cycle is the polymer-inherent isosterism of the carboxylates with the array of phosphates in nucleic acids and,