Báo cáo Y học: Inhibition of hyaluronan synthesis in Streptococcus equi FM100 by 4-methylumbelliferone doc

Inhibition of hyaluronan synthesis in Streptococcus equi FM100 by 4-methylumbelliferone Ikuko Kakizaki 1 , Keiichi Takagaki 1 , Yasufumi Endo 1 , Daisuke Kudo 1 , Hitoshi Ikeya 2 , Teruzo Miyoshi 2 , Bruce A. Baggenstoss 3 , Valarie L. Tlapak-Simmons 3 , Kshama Kumari 3 , Akio Nakane 4 , Paul H. Weigel 3 and Masahiko Endo 1 Departments of 1 Biochemistry and 4 Bacteriology, Hirosaki University School of Medicine, Hirosaki; 2 Research Center Denki Kagaku Kogyo Co. Ltd, Tokyo, Japan; 3 Department of Biochemistry and Molecular Biology, University of Oklahoma Health Sciences Center, Oklahoma City, Oklahoma, USA As observed previously in cultured human skin fibroblasts, a decrease of hyaluronan production was also observed in group C Streptococcus equi FM100 cells treated with 4-methylumbelliferone (MU), although there was no effect on their growth. In this study, the inhibition mechanism of hyaluronan synthesis by MU was examined using Strepto- coccus equi FM100, as a model. When MU was added to a reaction mixture containing the two sugar nucleotide donors and a membrane-rich fraction as an enzyme source in a cell- free hyaluronan synthesis experiment, there was no change in the production of hyaluronan. On the contrary, when MU was added to the culture medium of FM100 cells, hyaluro- nan production in the isolated membranes was decreased in a dose-dependent manner. However, when the effect of MU on the expression level of hyaluronan synthase was exam- ined, MU did not decrease either the mRNA level of the has operon containing the hyaluronan synthase gene or the protein level of hyaluronan synthase. Solubilization of the enzyme from membranes of MU-treated cells and addition of the exogenous phospholipid, cardiolipin, rescued hya- luronan synthase activity. In the mass spectrometric analysis of the membrane phospholipids from FM100 cells treated with MU, changes were observed in the distribution of only cardiolipin species but not of the other major phospholipid, PtdGro. These results suggest that MU treatment may cause a decrease in hyaluronan synthase activity by altering the lipid environment of membranes, especially the distribution of different cardiolipin species, surrounding hyaluronan synthase. Keywords: hyaluronan; synthesis; Streptococcus; 4-methyl- umbelliferone; phospholipids. Hyaluronan (HA) is a high molecular weight glycosamino- glycan composed of repeating disaccharide units of GlcNAc-b(1fi4)-GlcUA-b(1fi3) [1]. HA is one of the major components of the extracellular matrix together with proteoglycans and collagens, and is involved in many biological processes, including tissue organization, wound healing, tumor invasion and cancer metastasis, through its interactions with other extracellular matrix components [2,3]. It has long been suggested that HA may be implicated in malignant transformation and tumor progression [4]. There are many reports that HA production is increased in various tumor tissues including mesothelioma and Wilm’s tumor. Recently, a direct correlation between HA and tumorigenesis, and cancer metastasis was shown in studies using genetic manipulations to create mutant cells that were either overproducing HA or HA-deficient [5,6]. Overpro- duction of HA is also observed in diseases associated with inflammation and fibroses [3]. Many strains of group A and C Streptococci are able to synthesize HA [7,8]. Their thick HA coats surrounding the cell surfaces contribute to their pathogenicity by allowing them to escape from the immune systems of their hosts. The HA synthesized by Streptococci is not chemically or structurally distinguishable from that synthesized in mammalian cells. HA is synthesized by a membrane-associated hyaluronan synthase (HAS) from the precursors UDP-GlcUA and UDP-GlcNAc in either mammalian cells or Streptococci [9]. In the last several years, three distinct mammalian genes and three unique Streptococcal genes encoding the HASs have been cloned and their properties have been examined [9–12]. From the genomic analysis, it has been clarified that the has operon encodes for the HA synthesis system of Streptococci. The has operon is composed of three genes, hasA (which encodes the HA synthase), hasB (which encodes UDP- glucose dehydrogenase), and hasC (which encodes UDP- glucose pyrophosphorylase) [9]. Tlapak-Simmons et al.[13] demonstrated that the functional sizes of both the group A and the group C Streptococcus HASs are protein monomers in association with about 16 phospholipid molecules, in particular cardiolipin (CL), which was also shown to be necessary for optimal enzymatic activity [14]. Due to the Correspondence to M. Endo, Department of Biochemistry, Hirosaki University School of Medicine, 5 Zaifu-cho, Hirosaki 036–8562, Japan. Fax: + 81 172 39 5016, Tel.: + 81 172 39 5015, E-mail: endo-m@cc.hirosaki-u.ac.jp Abbreviations: CL, cardiolipin; DDM, n-dodecyl-b- D -maltoside; GlcNAc, N-acetylglucosamine; GlcUA, glucuronic acid; HA, hyaluronan (hyaluronic acid); HABP, hyaluronan binding protein; HAS (Has), hyaluronan synthase; MU, 4-methylumbelliferone; spHAS, S. pyogenes HAS; seHAS, S. equisimilis HAS. Note: A web site is available at http://www.med.hirosaki-u.ac.jp/ bioche1/test/Biochem-top/Biochem-top1.html (Received 20 June 2002, revised 14 August 2002, accepted 29 August 2002) Eur. J. Biochem. 269, 5066–5075 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.03217.x cloning of the HAS genes, it has been possible to genetically manipulate the production of HA, and consequently, correlations between HA production and various biological processes have now been brought to light. For example, the effects of antisense inhibition of HA production on the organization of the extracellular matrix in human articular chondrocytes has been examined [15]. Studies using targeted deletion of HAS genes have also been made to investigate theroleofHAin vivo.ItwasreportedthatHas2 + embryos, which lack HA production by Has2, exhibit severe cardiac and vascular abnormalities and die during fetal develop- ment [16]. However the details about the multiple functions of HA have not been fully established. We found that HA synthesis in cultured human skin fibroblasts was inhibited by 4-methylumbelliferone (MU, 7-hydroxy-4-methyl-2H-1-benzopyran-2-one) with no effect on the synthesis of any other glycosaminoglycan and that an HA-deficient extracellular matrix was formed [17,18]. Some agents have been reported to inhibit HA synthesis, however, no clearly specific inhibitors for HA synthesis have been found [19–23]. Although its biochemical mechanism of inhibition is not well understood, MU has been used in some studies on the function of HA [5,24,25]. For example, it was used to prepare an HA-deficient transfected cell line expressing a HAS gene in order to examine the role of HA in tumorigenesis [5]. Recently, Endo et al. investigated the correlation between HA and the other components of extracellular matrices, using cultured human skin fibroblasts in which HA production was inhibited by MU treatment [24]. In order to elucidate the inhibition mechanism of HA synthesis, in the present study we have examined the effect of MU on prokaryotic cells, S. equi FM100, as a model. We find that, as in cultured human skin fibroblasts [18], MU did not directly inhibit HAS in vitro but did inhibit the enzymatic activity of HAS in intact cells. Furthermore, we show that MU does not directly inhibit the processes of transcription or translation of HAS, but that a possible novel mechanism of inhibition of HA synthesis by MU is probably due to an alteration of the lipid environment of the Streptococcal membranes. MATERIALS AND METHODS Materials Lysozyme and MU were purchased from Wako Pure Chemicals (Osaka, Japan). MU was dissolved in dimethyl- sulfoxide, and the final concentration of dimethylsulfoxide in the culture medium and reaction mixtures for HA syn- thesis was 0.1%. UDP-[U- 14 C]GlcUA (270 mCiÆmmol )1 ) was purchased from American Radiolabeled Chemicals (St. Louis, MO, USA). UDP-GlcNAc, ATP, dithiothreitol, bovine testicular hyaluronidase and bovine heart CL were purchased from Sigma (St. Louis, MO, USA). HA from human umbilical cords and Streptomyces hyaluronidase were obtained from Seikagaku Corporation (Tokyo, Japan). Actinase E was from Kaken Pharmaceutical (Tokyo, Japan). Hyaluronic Acid ÔChugaiÕ quantitative test kit for the sandwich binding protein assay was purchased from Chugai Pharmaceutical (Tokyo, Japan) [26]. The random-primed DNA labeling kit was from Amersham Pharmacia Biotech (Tokyo, Japan) and [a- 32 P] dCTP was from NEN Life Science Products (Boston, MA, USA). Antiserum raised against the whole HAS from Streptococcus pyogenes (spHAS) was described previously [10,27]. Anti- rabbit Ig conjugated to horseradish peroxidase was from Dako Japan (Kyoto, Japan) and n-dodecyl-b- D -maltoside (DDM) was from Nakarai Tesque (Kyoto, Japan). Culture of Streptococci and treatment with MU Encapsulated group C S. equi FM100 was derived from S. equi (ATCC9527) and maintained at 33 °C in a synthetic solid medium (pH 8), which we have modified from the medium reported by van de Rijn and Kessler [28]. Cells were precultured in liquid medium (1.5% polypeptone-S/0.5% yeast extract/0.2% dipotassium hydrogenphosphate/0.16% sodium thiosulfate/0.02% sodium sulfate/2.0% glucose, pH 8), and then added to 100 volumes of fresh medium and grown with or without MU. Cell numbers were estimated by measuring the absorbance at a wavelength of 660 nm. Cells were stained with nigrosin and observed through a microscope (OLYMPUS, model BHS). The detailed infor- mation for the generation and characterization of S. equi FM100 is described in Japanese patent (JP1750524, Japan Patent Office). Analysis of the HA released into the culture medium Exponentially growing cells were cultured with or without various concentrations of MU (0.2, 0.5, 1.0 and 2.0 m M ). At the various time points (0, 3, 5.5, 8 and 22 h), HA released into the culture medium was measured by the sandwich binding protein assay using hyaluronan binding protein (HABP) according to the manufacture’s instructions for the hyaluronic acid ÔChugaiÕ quantitative test kit [26]. The molecular size of HA released into the culture medium was analyzed by gel filtration HPLC using a Shodex OHpak KB-805 column (8 · 300 mm). Elution was with 0.2 M NaCl at a flow rate of 0.5 mLÆmin )1 .Eluted fractions were monitored by detecting absorbance at a wavelength of 215 nm. The molecular sizes of standard HA were 1.0, 3.0, 4.1, 8.0, 12 and 19 · 10 5 [29]. Preparation of a membrane-rich fraction and solubilization The membrane-rich fraction was prepared by following the method of Sugahara et al. [30]. Briefly, exponential phase cell cultures were harvested by centrifugation at 18 000 g for 30 min. Then, cells were suspended in 0.05 M sodium and potassium phosphate buffer, pH 7.4, containing 5 m M dithiothreitol. The cell suspension was sonicated on ice using a Branson Sonifer (model 250) for 1 min. The disrupted cells were centrifuged at 10 000 g for 10 min and the supernatant fluid was withdrawn. Following centrifugation of the 10 000 g supernatant fluid at 105 000 g for 60 min, the resultant pellet was washed with fresh buffer by centrifugation at 229 000 g for 45 min. The pellet was suspended in 0.033 M sodium and potassium phosphate buffer, pH 7.4 containing 5 m M dithiothreitol, and used as an enzyme source for the cell-free HA synthesis assay. Solubilization of membranes using DDM was performed as previously reported by Tlapak-Simmons et al. [14]. Each cell-free assay was standardized for the amount of protein used in order to compare the activities between the Ó FEBS 2002 Inhibition of HA synthesis in Streptococcus by MU (Eur. J. Biochem. 269) 5067 whole and partially solubilized membranes. Protein content was determined by the method of Bradford [31], and the results were confirmed by SDS/PAGE analysis of all the samples, according to the method of Laemmli [32]. Assay of cell-free HA synthesis Cell-free HA synthesis was performed by a modification of the method reported by Nakamura et al. [18]. Analysis of the transfer of UDP-[U- 14 C] GlcUA to HA was monitored as follows. Assay mixtures contained the following compo- nents in a final volume of 0.1 mL: 47 m M sodium and potassium phosphate buffer (pH 7.1), 5 m M dithiothreitol, 5m M MgCl 2 ,100l M UDP-GlcNAc, 3.7 l M UDP-[U- 14 C] GlcUA (1.8 lCi), 5 m M ATP, and 12.5–250 lgofmem- brane-rich fractions or DDM extracts as the enzyme source. MU was added as indicated to the assay mixture or the culture medium. For certain experiments, the enzyme was preincubated with 2 m M bovine heart CL. The assay mixture was incubated at 37 °Cfor1.5h,andthereaction was terminated by boiling for 2 min. Because the reaction was linear up to 1.5 h of incubation time, this time point wasused.AfteractinaseEdigestion(2.5mgÆmL )1 ,16h at 45 °C), 1/6 volume of 50% trichloroacetic acid was added with mixing, the reaction mixture was cooled on ice, and the supernatant was withdrawn after centrifugation. In the presence of 50 lg of carrier HA, ethanol precipitation was performed five times to remove the unincoporated, free radioisotopes. The final precipitate was dissolved in water, digested with Streptomyces hyaluronidase and then precipitated with ethanol. The radioactivity remaining in the supernatant following ethanol precipitation, which represents the digestion products specifically derived from HA, was then determined using a liquid scintillation counter. Isolation of S. equi FM100 hasA gene To obtain a probe for hybridization, the 1166 bp region of the S. equi FM100 hasA gene was amplified with primers based upon the nucleotide sequence of the group C Streptococcus equisimilis hasA gene (seHAS gene, GenBank Accession Number AF023876) [10], using genomic DNA from FM100 cells as a template. Genomic DNA was extracted following the method of Sambrook et al.[33]after digestion with lysozyme and bovine testicular hyaluroni- dase. The primer set used in the PCR had the following sequence: forward, 5¢-ACTGTTGTGGCCTTTAGTA-3¢ and reverse, 5¢-AAGGGCTGTAGGACAAACAA-3¢.The sequence of the amplified product completely matched the region of nucleotides 25–1190 of AF023876. RNA preparations and Northern hybridization Total RNA was prepared from FM100 cells using a RNeasy Plant Mini Kit (Qiagen Japan, Tokyo, Japan) according to the manufacturer’s specifications. Five micro- grams of RNA, denatured with formaldehyde and forma- mide, was separated on a 1% (w/v) agarose gel containing 1.1 M formaldehyde and transferred to a nylon membrane (Hybond-N, Amersham, Buckinghamshire, UK) in 20· NaCl/Cit (3.0 M NaCl/0.3 M sodium citrate, pH 7.0). The hasA DNA probe was labeled by the random priming method [34]. Hybridization was performed with a 32 P-labeled probe at 42 °C for 18 h in 50% formamide, 3· NaCl/Cit, 0.05 M Tris/HCl (pH 7.5), 1 m M EDTA, 0.02% BSA, 0.02% Ficoll, 0.02% polyvinylpyrrolidone, 20 lgÆmL )1 tRNA, 20 lgÆmL )1 herring sperm DNA. Then the filters were washed twice with 3· NaCl/Cit, 0.1% SDS at 37 °C for 30 min, and twice with 0.1· NaCl/Cit, 0.1% SDS at 50–65 °C for 30 min. Autoradiography was carried out by exposure to X-ray film (Kodak X-Omat AR) at )80 °C using an intensifying screen. Results of autoradio- graphy were quantified using NIH IMAGE (version 1.62) software. SDS/PAGE and immunoblotting SDS/PAGE was performed in 10% acrylamide gels by the method of Laemmli [32]. Protein was stained with the Coomassie brilliant blue R-250. For immunoblotting, proteins were transferred to a polyvinylidene fluoride filter (Millipore Japan, Tokyo, Japan), and stained with anti-spHAS antibody according to the method of Towbin et al.[35]using3,3¢-diaminobenzidine tetrahydrochloride (Dojindo Laboratories, Kumamoto, Japan) for detection. Resulting bands were quantified using NIH IMAGE (version 1.62) software. Extraction of lipids Lipids were extracted using the method of Folch et al.with minor modifications [36]. Membrane samples were dis- persed in 100 lL of chloroform/methanol (2 : 1, v/v) per milligram membranes by brief sonication and then shaken gently for 40 min at room temperature. The sample was then centrifuged at 3800 g for 10 min and the supernatant fluid was drawn off, with care not to remove any pellet material. A 0.2 · the volume of 0.9% NaCl was added to the sample, which was mixed vigorously by vortexing twice for 30 s. The mixture was centrifuged at 420 g and the top aqueous layer was removed. The bottom chloroform layer containing the phospholipids was analyzed directly or saturatedwithnitrogenandstoredat)20 °C. Mass spectrometric analysis The MALDI-TOF mass spectrometer used was a Voyager Elite (Applied Biosystems, Framingham, MA, USA) equipped with an N 2 laser (337 nm) located in the NSF EPSCoR Oklahoma Laser MS Facility (OUHSC). Samples were analyzed in the reflector, negative ion mode using a delayed extraction of 200 ns, a grid voltage of 79%, and were subjected to a 20-kV accelerating voltage. An external calibration was obtained using bovine heart CL, which has a mass of 1448.97. The matrices used were 6-aza-2- thiothymine or 2,4,6-trihydroxyacetophenone at 5 mgÆmL )1 in chloroform/methanol (2 : 1, v/v) containing 10 m M dibasic ammonium citrate. Samples were diluted with two volumes of chloroform/methanol (2 : 1, v/v) and then mixed 1 : 1 with the matrix solution prior to spotting on a sample plate and air drying. Spectra are an average of 80–100 scans. In some cases the identity of specific m/z species was confirmed by post source decay analysis in both the positive and negative ion modes. Total amounts of CL or PtdGro recovered from FM100 cells were assessed based on their 5068 I. Kakizaki et al. (Eur. J. Biochem. 269) Ó FEBS 2002 signal intensities relative to appropriate phospholipids that were used as standards. RESULTS Inhibition of HA production of FM100 cells To examine the effect of MU on the growth of FM100 cells, the cells were cultured in liquid medium with or without MU for various periods, and cell numbers were estimated by measuring absorbance at 660 nm at each time point. No significant effect on the growth was observed in the range of 0–2.0 m M MU (data not shown). Absorbance at 660 nm was measured in all subsequent experiments, however, inhibition of growth of FM100 cells by MU was not observed. On microscopic examination, untreated control cells formed HA coats on their cell surfaces, and the coat grew thicker over the course of culture time (Fig. 1). However the formation of the HA coating was dose-dependently decreased, when the cells were cultured with 0.2–2.0 m M MU. The decrease in coat formation was observed as soon as 3 h after culturing in the presence of 0.2 m M MU(data not shown), and became more marked after longer incuba- tion times. When FM100 cells were cultured for 22 h, the HA coats on cell surfaces essentially disappeared (Fig. 2A). Only a very thin HA coating was observed at the highest concen- tration of MU in the 22 h cultures (data not shown). On the other hand, release of HA into the culture medium, as assessed by HPLC analysis, was observed after 8 h, and increased with continued incubation reaching a peak after 36 h (data not shown). The size distribution of HA released into the culture medium was then analyzed by HPLC, after the cells were cultured with various concentrations of MU (0–2.0 m M ) for 22 h (Fig. 2B). Regardless of the MU concentration, the molecular sizes of the major peaks of HA did not shift and were calculated at 1.2–1.9 · 10 6 .These peaks disappeared after digestion with the very specific Fig. 1. Effect of MU treatment on the HA coat formation of S. equi FM100 cells. Microscopic photograph of HA coats (arrowheads) on cell surfaces of FM100. A–C, untreated cultures; D–F, treated with 0.5 m M MU; G–I, with 1.0 m M MU. A, D and G, cultured for 3 h; B, E and H, cultured for 5.5 h; C, F and I, cultured for 8 h. Original magnification, · 1000. Magnification bar represents 10 lm. Fig. 2. Analysis of HA released into the culture medium. FM100 cells were cultured with various concentrations of MU (0–2.0 m M )for22 h. (A) Micrograph of the HA coats on cell surfaces cultured for 8 h (a) andfor22h(b)withoutMU.(B)HPLCanalysisofHAreleasedinto the culture medium. A Shodex OHpak KB-805 column (8 · 300 mm) was used and eluted with 0.2 M NaCl at a flow rate of 0.5 mLÆmin )1 . Eluted fractions were monitored at a wavelength of 215 nm. Arrow- heads indicate the peak of HA. Ó FEBS 2002 Inhibition of HA synthesis in Streptococcus by MU (Eur. J. Biochem. 269) 5069 Streptomyces hyaluronidase (data not shown). However, when the secretion of HA into the culture medium by FM100 cells was quantified by measuring the HA peak areas, it was found that HA production was clearly decreased by MU. To verify this effect and study it further, FM100 cells were cultured in the presence of MU for various periods (0, 3, 5.5, 8 and 22 h), and the HA production in the culture medium was quantitated by a sandwich binding protein assay using a very specific HABP (Fig. 3). By 8 h of culturing, a small amount of HA released into the culture medium was observed, but at longer times HA production was markedly increased. After 22 h of culturing, control cells treated with only dimethylsulfoxide produced 725 ngÆmL )1 of HA. However, HA production and accumulation in the medium was dose-dependently decreased by MU treatment. The HA production by the cells treated with 1.0 m M MUwasdecreasedto 300 ngÆmL )1 , about 40% of the control value. No signifi- cant HA production was detected at a concentration of 2.0 m M . These inhibition effects by MU were reversed by washing the cells after 22 h of MU treatment, resuspending in fresh medium and allowing them to grow for 14 h without or with diluting the cells (data not shown). The effect of several analogues of MU on the HA synthesis in FM100 cells was also examined. The inhibition of HA production was not observed except in the case of the sodium salt of MU (MU-Na), although the extent of inhibition was less than with MU (data not shown). Effect of MU on cell-free HA synthesis In order to examine whether the addition of MU to a membrane-rich fraction could inhibit HAS activity, a cell-free HA synthesis experiment was performed. A mem- brane-rich fraction was prepared from cultured FM100 cells by sonication and ultracentrifugation, and used as an enzyme source. UDP-[U- 14 C] GlcUA and UDP-GlcNAc were used as donors, and the transfer of UDP-[U- 14 C] GlcUA to newly synthesized HA was analyzed. The activity in the membrane-rich fraction was hardly inhibited by MU up to 1.0 m M (data not shown). This result suggest that the inhibition of HA production was not caused by direct inhibition of HAS activity. Effect of MU on HAS activity in FM100 cells treated with MU The activity of the HAS in FM100 cells cultured with various concentrations of MU for 12 h was also measured. Membrane-rich fractions were prepared, and their ability to support cell-free HA synthesis was determined. HA pro- duction by these isolated membranes was decreased by MU treatment of the live cells in a dose-dependent manner (data not shown). At 2.0 m M , HA production was decreased to about 10% of control value. Effect of MU treatment on HAS expression level in FM100 cells To examine whether MU inhibits the expression of the has operon, northern hybridization was performed using S. equi FM100 hasA DNA as a probe. As has operon mRNA was reported to be detected only at the exponential phase of growth [37], which we also observed in a preliminary experiment, a 4.5-h culture was used. The probe hybridized to a 4.1-kb mRNA, corresponding to the has operon (Fig. 4). Although treatment with 0.2 m M MUfor3h resulted in inhibition of HA coat formation (data not shown), has operon mRNA levels were hardly affected by up to 2.0 m M MU. The protein level of HAS was also examined after the FM100 cells were incubated with MU (Fig. 5). Fig. 4. Effect of MU on has operon mRNA level in S. equi FM100. Total RNA was extracted from FM100 cells that had been treated with various concentrations of MU (0–2.0 m M ) for 4.5 h. Then, the level of has operon mRNA was analyzed by northern hybridization using S. equi FM100 hasA DNA as a probe. Lanes 1, 2, 3, 4 and 5 contained mRNA from FM100 cells treated with 0, 0.2, 0.5, 1.0 and 2.0 m M MU, respectively. Each lane contained 5 lgoftotalRNAandtheRNA staining pattern is shown at the bottom. Fig. 3. Effect of MU treatment of S. equi FM100 cells on HA accu- mulation in culture medium. FM100 cells were cultured with or without MU for various periods (0, 3, 5.5, 8 and 22 h). HA production and release into culture medium of the FM100 cells was quantitated by the sandwich binding protein assay. Time represents hours after addition of a 1/100 volume of inoculum into fresh medium. Symbols are as follows; d,0;h,0.2m M ; j,0.5m M ; n,1.0m M ; m,2.0m M MU. 5070 I. Kakizaki et al. (Eur. J. Biochem. 269) Ó FEBS 2002 The anti-spHAS antibody used here is also strongly cross-reactive with the seHAS, the HAS from group C S. equisimilis [10]. In this experiment the antibody recog- nized a protein of M r 48 000, the size expected for the calculated relative molecular mass (M r 47 778) of seHAS [10] (Fig. 5A). The HAS protein level, however, did not change by treatment of cells with up to 1.0 m M MU. At 2.0 m M , however, a decrease in the level of total protein containing HAS was observed (about 20% of the control value of HAS content remained). These results are repro- ducible. Two additional weak bands, which migrated at higher molecular weights than HAS were also observed. These appear to be nonspecific bands, unrelated to HAS protein. Effects of MU treatment on the phospholipid composition of membranes Streptococcal HAS, solubilized from membranes, is dependent on the presence of exogenous phospholipids, particularly CL, for optimal enzyme activity [14,38]. There- fore, we examined the possibility that MU treatment of FM100 cells results in a modified composition of membrane phospholipids, which might in turn alter the environment of lipids that surround and interact with the HAS. Membranes were solubilized with DDM, and exogenous CL was added tothereactionmixtureinthecellfreesystembeforethe substrates were added (Fig. 6). When whole membranes were preincubated with exo- genous CL, the HAS activity was not affected. However, when membranes were solubilized with DDM an increase in HAS activity was then observed in each preparation of membranes from cells treated with up to 1.0 m M MU (Fig. 6, striped bars). The increased rate of HA synthesis after solubilization with DDM was about 30% of the activ- ity in whole membranes. The enhancement of HAS activity when membranes are solubilized by DDM has already been reported [14]. Further stimulation of HAS activity by addition of exogenous CL to the DDM extract resulted in recovery of HAS activity to about the level in samples not treated with MU (Fig. 6, solid bars). Even when cells were treatedwith2.0m M MU, subsequent stimulation of HAS by CL was observed in DDM extracts, although the stimulated activity did not reach that of MU-untreated samples. This deficiency of reconstituting HA synthesis activity at 2 m M MU can be explained by the decreased level of HAS protein noted above. CL addition did not cause a significant increase in HAS activity in the extracts from untreated cells. A key question in the experiment of Fig. 6 is whether there might be a differential solubilization of some CL species by the detergent DDM, so that the membrane composition of CL is not reflected in the DDM extract. To address this issue, Folch extractions were performed on untreated membrane preparations and on the DDM Fig. 6. Effects of detergent solubilization and addition of cardiolipin on HAS activities in membranes from MU-treated cells. FM100 cells were cultured with or without various concentrations of MU (0–2.0 m M )for 12 h. Then, the effect of solubilization with DDM and addition of 2.0 m M of exogenous cardiolipin (CL) on HAS activity of membranes was examined in the cell-free system. Columns are as follows: unshaded bars, whole membrane, CL(–); dotted bars, whole mem- brane, CL(+); striped bars, solubilized membrane, CL(–); solid bars, solubilized membrane, CL(+). Data shown are the mean of triplicate assays and the bars represent SD. Fig. 5. Effect of MU on HAS protein level in S. equi FM100. Mem- brane-rich fractions were prepared from FM100 cells treated with various concentrations of MU (0–2.0 m M ) for 12 h. Then, they were analyzed for HAS protein by SDS/PAGE and immunoblotting with anti-spHAS antibody. (A) Immunoblotting. (B) Coomassie brilliant blue R-250 staining pattern. Lanes 1, 2, 3, 4 and 5 in A and B contained membrane-rich fractions (10 lg protein) derived from FM100 cells treated with 0, 0.2, 0.5, 1.0 and 2.0 m M MU, respectively. Ó FEBS 2002 Inhibition of HA synthesis in Streptococcus by MU (Eur. J. Biochem. 269) 5071 extracts made from membranes treated with DDM. The samples were processed identically and analyzed by MALDI-TOF MS as described in the Experimental proce- dures. The average signals (peak heights) at a given CL mass were calculated for a series of three samples, each analyzed in duplicate (n ¼ 6). Within a given sample, the mass signals for various CL species (e.g. as shown in Fig. 8, bottom panel) were arbitrarily normalized to one of the largest peaks, which was given a value of 1.0. The other mean relative peak heights in each Folch extract typically varied from 0.2 to 0.7 and their standard errors were ±5–15% of these values. There were no statistically significant differ- ences between the two Folch extracts for any of the CL species detected (not shown). The pattern and relative intensities of CL species was the same in membranes or in the DDM extract derived from those membranes. MALDI-TOF mass spectrometric analysis was then performed to determine whether the phospholipid profile of cell membranes was altered by MU treatment. Phos- pholipids were extracted from the membranes of FM100 cells cultured with 0–2.0 m M MU for various times, and the lipids present in the extracts were then analyzed. No change in the total amount of CL was observed after MU treatment (data not shown). Only two major classes of phospholipids were detected by MALDI-TOF MS analysis of Folch extracts, CL and PtdGro. Other phospholipids including PtdEtn, PtdCho, PtdSer and PtdIns were not detected. As reported by others, Gram-positive bacteria such as S. equi typically have essentially only these two phospholipids [39]. Although only two phospholipids were present, their diversity was striking, as at least 60 variants of CL and PtdGro could be identified. The pattern and relative abundance of the minor and major PtdGro species was not altered by treatment with MU (data not shown). Unexpectedly, there were no obvious or reproducible changes in the composition of the multiple CL species, when cells were treated with MU. One of these experiments is shown in Fig. 7 for FM100 cells cultured with or without MU for 12 h. The observed CL pattern (Fig. 7, bottom panel) is a cluster of ‡5–7 m/z species, as each CL has four fatty acyl chains, each of which can have a different number of double bonds (e.g. 0–3) and carbons (e.g. C16–C20). At least nine clusters of peaks were identified as CL species, based on their characteristic m/z pattern [14] and, in some cases, the specific fragments obtained by post source decay analysis (not shown). These multiple clusters of CL species were designated by the letters A–I. The same CL species observed in the control (0 m M MU) were also observed in membranes from the MU-treated cells. None- theless, there was a potentially interesting effect of MU on the distribution of CL species present in cells treated with MU for increasing times. We observed in multiple spectra that the relative amounts of CL species appeared to change in treated cells. To assess this possibility more quantitatively, the m/z signals for all the CL species, A–I, were integrated and the percent of the total area represented by each CL species was then calculated (Table 1). Exposure of cells to MU caused a reproducible decrease in the relative amount of the smaller mass CL species and a corresponding increase in the larger mass CL species. Similar results were also obtained with a second matrix molecule, 2,4,6-trihydroxy- acetophenone. The distribution pattern of bovine heart CL Fig. 7. MALDI-TOF mass spectrometric analysis of cardiolipin in the membranes of MU-treated S. equi FM100. Phospholipids were extracted from the isolated membranes of FM100 cells, which had been cultured without or with 1.0 or 2.0 m M MU for 12 h. The phospholipid profile in the CL mass region was then analyzed by MALDI-TOF MS. The matrix used was 6-aza-2-thiothymine. Multiple CL peaks, which are labeled A through I, were detected. The distribution of these multiple CL species (as a percent of the total CL) is summarized in Table 1. Minor peaks that differ from a more major peak by 1 m/z unit represent mass variants due to natural isotope abundance (e.g. one more 13 C present in the molecule instead of 12 C as in a neighboring peak). 5072 I. Kakizaki et al. (Eur. J. Biochem. 269) Ó FEBS 2002 added in the cell-free HA synthesis experiment (Fig. 6) was also analyzed, and was compared with that of the CL in the FM100 cells. The bovine heart CL has the same compo- sition as that reported in a previous paper (14). It contained one major species with m/z-value of 1448.97. This bovine CL species is relatively large and, although it may not be present in FM100 cells, is intermediate in mass between two of the major CL species (i.e. peaks E and F in Fig. 7). DISCUSSION As shown previously in fibroblasts, MU also did not affect the molecular size distribution of HA produced by FM100 cells. This suggests that MU acts to inhibit the HA synthesis pathway but not to stimulate the HA degradation pathway. In the present study, we showed that MU did not directly inhibit HAS activity even at a relatively high concentration, and this result agrees with that obtained using cultured human skin fibroblasts. It seems likely therefore that the mechanisms of inhibition of HA synthesis by MU are very similar if not identical between eukaryote and prokaryote cells. Cell-free experiments also showed that the decrease of HA production by MU is not due to a decrease in the intracellular concentration of the sugar-nucleotide precur- sors, UDP-GlcUA and UDP-GlcNAc. Although the reac- tion mixtures contained a very large molar excess of these donors, well above their K m values [40], decreased HA production was nonetheless still observed in the membrane- rich fraction from the cells preincubated with MU. Furthermore, the level either of transcription or translation of HAS was hardly affected by MU. As Western analysis revealed similar amounts of the HAS protein in the MU-treated membranes, except at 2 m M , the delivery of HAS to the Streptococcal cell membrane was not inhibited byMUupto1m M . Because the level of total protein decreased at 2 m M MU, MU must have nonspecific effects on various proteins at very high concentration. Addition of MU to a membrane-rich fraction did not inhibit its HAS activity, whereas addition of MU to live cells did inhibit the HAS activity of their membranes. In order to examine whether possible metabolites of MU are involved in the inhibition of HA synthesis in intact cells, we performed some experiments using HPLC or ion-chromatography. However, we did not detect any metabolites of MU either in the culture supernatant or in the cell extract from FM100 cells cultured with MU (data not shown). We believe therefore there is no or little involvement of MU-meta- bolites in the inhibition of HA synthesis by MU in FM100 cells. These above results indicate that something required for a fully functional HA synthesis system is down-regulated or inhibited in intact cells exposed to MU. The presence of post-translational modifications in the native enzyme in Streptococci has not been addressed. Thus, it is likely that MU may alter either a required event needed to generate active HAS enzyme or the availability of a required activator such as CL or some other event needed to generate active enzyme. It has been suggested by other investigators that HA synthesis in mammalian cells and Streptococcal cells is strictly controlled by complicated mechanisms, including phosphoryl modification of HAS [21,41–43]. It should also be noted that multiple consensus sequences for phosphory- lation by some kinases are found in HASs [44,45]. Based on protein motif analysis using the PROSITE database (release 16.22), multiple potential phosphorylation sites could also be found in S. equi FM100 HAS. However, no change in the phosphorylation-level of HAS protein by MU-treat- ment was observed when it was examined by Western analysis using anti-phosphoamino acid antibodies (data not shown). Furthermore, previous MALDI-TOF analysis of purified recombinant Streptococcal HAS demonstrated that the enzyme contains no stoichiometric covalent modifica- tions [13]. Thus, we conclude that there is no involvement of phosphoryl control for the inhibition of HA synthesis by MU. The HAS is a transmembrane protein, and it has been suggested that the monomer Streptococcal HAS forms a pore-like structure with 14–18 molecules of CL, as an active enzyme [13,14]. It has also been suggested that the HA chain polymerized at the inner surface of the plasma membrane is translocated to the outside of the cell through this intrinsic enzyme pore [14]. Recently, the first topological organiza- tion of the spHAS protein was determined experimentally, not only by algorithms, and the requirement of lipid association for the formation of the pore and for the Table 1. Effect of MU on the distribution of CL in FM100 cells. Lipids were extracted from the isolated membranes of FM100 cells, which were cultured without or with 1.0 or 2.0 m M MU for 12 h, and analysed by MALDI-TOF MS. Percent distribution of CL species (Fig. 7,A–I) was summarized. The matrix used was 6-aza-2-thiothymine. Each value represents the mean of triplicate spectra ± SEM. MU(m M ) 0 1.0 2.0 Peak number Average % SEM Average % SEM Average % SEM A 4.52 0.72 3.36 0.22 3.12 1.48 B 6.52 0.43 4.85 0.83 3.41 0.52 C 9.59 0.51 7.74 0.98 6.04 0.67 D 14.61 0.20 11.27 0.28 10.03 0.84 E 20.59 1.50 20.02 0.98 18.91 1.45 F 18.78 1.38 26.37 0.99 29.30 1.09 G 9.56 0.86 12.44 0.18 14.39 0.14 H 7.91 0.28 7.64 0.50 8.06 0.88 I 7.92 0.97 6.33 0.41 6.65 1.00 Ó FEBS 2002 Inhibition of HA synthesis in Streptococcus by MU (Eur. J. Biochem. 269) 5073 stabilization of the enzymatic activity was again suggested [46]. If this pore model is correct, then another possibility for how MU inhibits HA synthesis is that this HA translocation process may be blocked by MU through subtle changes in the steric conformation of the HAS protein or the mem- brane bilayer. Because MU is very lipid soluble the membrane-bound HAS or the organization of lipids in the membrane itself may be very sensitive to this compound. Alternatively, the glycosyltransferase activities of HAS or its HA translocation activity, all of which are very dependent on the proper conformation of this membrane protein, may be adversely affected by MU in an indirect way. Our results suggest that a possible mechanism of inhibition is that MU alters the phospholipid distribution of the cell membrane, which could then destabilize the HAS activity. MALDI-TOF mass spectrometric analysis indi- cates that FM100 cells contain predominantly only PtdGro and CL as their major phopholipids. In fact, it may not be a coincidence that CL is a major membrane lipid, as it is required by HAS. Natural selection of cells able to synthesize large HA coats may have resulted in a compo- sition of membrane phopholipids compatible with high HAS activity. A key finding in the present study is that HAS inhibition, in membranes isolated from MU-treated cells, is rescued by solubilizing the enzyme in DDM and then providing endogenous CL. Even without MU treatment, HAS activity is higher in DDM-solubilized membranes than in whole membranes. Although we do not have a complete explanation for this latter effect, it is not unusual to find enhanced activity of membrane-bound enzymes after detergent solubilization. The finding here is very reprodu- cible and is consistent with the same observation made in an earlier paper by one of our groups reporting the purification of HAS [14]. This earlier study found that group C HAS activity was enhanced  20% by solubilization of the membranes in DDM and in the present study, using a different group C strain, the stimulation was  30%. One explanation for the enhancement with DDM may simply be that the rate of HA synthesis by the DDM- solubilized enzyme is not as diffusion-controlled in its ability to encounter and utilize the substrates, as it might be when membrane-bound. Another possibility is that DDM micelles may reconstitute a membrane-like environment in which the enzyme is intrinsically more active. For example, the HA translocation function, in which the growing HA chain traverses the bilayer in an intact membrane, may be more efficient in the more flexible artificial environment of micelles, thus enabling the enzyme to be less hindered and to polymerize HA at a faster rate. The ability of exogenous CL to rescue inhibited, DDM- solubilized HAS suggests that the enzyme in the membranes of MU-treated cells, is inactive because it is unable to interact with CL species that are able to activate it more optimally. The complexity of the natural pattern of CL in FM100 cells is very impressive with well over 50 discrete, identifiable species. In this regard, our most interesting result indicated that with increasing time of MU treatment there was a decrease in the proportion of smaller CL species and an increase in the larger species. Although we do not know exactly what this observation means for the activity of HAS in membranes of MU-treated live cells, the results provide a possible explanation for the inhibition of HAS by MU and the rescue of solubilized inhibited HAS by CL, because the enzyme is lipid-dependent and relatively CL-specific for its activity. Our interpretation at this point is that in order to be optimally active, the HAS may require or prefer to interact with CL species containing fatty acids with a particular chain length and unsaturation pattern, and that the MU treatment of cells decreases the availability of these favorable CL species. Additionally, the interaction of HAS with CL species that have quite different fatty acid components (e.g. larger or with a different number and location of double bonds) may actually inhibit the enzyme, so that in live cells the HAS activity could be decreased by MU treatment as the cellular distribution of CL species changed. The isolated membranes from the cells treated with MU show a similar inhibition of HAS activity because the enzyme is still associated with these ÔbadÕ CL species. However, when these membranes are solubilized and exogenous CL is added, the enzyme can then interact with the CL species it prefers and become reactivated. Further study will be required to confirm this interpretation and to understand fully the mechanisms for inhibition of HA synthesis by MU. This information may be useful in the treatment of diseases involving excess production of HA. ACKNOWLEDGMENTS This work was supported by Grants-in Aid (Nos. 08457032, 09240202, 09358013, 11476029, 12680603 and 12793010) for Scientific Research from the Ministry of Education, Culture, Sports, Science, and Tech- nology of Japan and by National Institutes of Health grant GM35978 from the National Institute for General Medical Sciences, USA. REFERENCES 1. Weissman, B. & Meyer, K. (1954) The structure of hyalobiuronic acid and of hyaluronic acid from umbilical cord. J. Am. Chem. Soc. 76, 1753–1757. 2. Laurent, T.C. & Fraser, J.R.E. (1992) Hyaluronan. FASEB J. 6, 2397–2404. 3. Knudson, C.B. & Knudson, W. (1993) Hyaluronan-binding pro- teins in development, tissue homeostasis, and disease. FASEB J. 7, 1233–1241. 4. Knudson,W.,Biswas,C.,Li,X.Q.,Nemec,R.E.&Toole,B.P. (1989) The role and regulation of tumour-associated hyaluronan. CIBA Found. Symp 13, 150–159. 5. Kosaki, R., Watanabe, K. & Yamaguchi, Y. (1999) Over- production of hyaluronan by expression of the hyaluronan synthase Has2 enhances anchorage-independent growth and tumorigenicity. Cancer Res. 59, 1141–1145. 6. Itano, N., Sawai, T., Miyaishi, O. & Kimata, K. (1999) Relationship between hyaluronan production and metastatic potential of mouse mammary carcinoma cells. Cancer Res. 59, 2499–2504. 7. Krause, R.M. (1972) The Streptococcal cell: relationship of structure to function and pathogenesis. In Streptococci and Streptococcal Diseases (Wannamaker, L.W. & Matsen, J.M., eds), pp. 3–18. Academic Press, New York. 8. Wessels, M.R., Moses, A.E., Goldberg, J.B. & DiCesare, T.J. (1991) Hyaluronic acid capsule is a virulence factor for mucoid group A Streptococci. Proc. Natl Acad. Sci. USA 88, 8317–8321. 9. Weigel, P.H., Hascall, V.C. & Tammi, M. (1997) Hyaluronan synthases. J. Biol. Chem. 272, 13997–14000. 10. Kumari, K. & Weigel, P.H. (1997) Molecular cloning, expression, and characterization of the authentic hyaluronan synthase from group C Streptococcus equisimilis. J. Biol. Chem. 272, 32539– 32546. 5074 I. Kakizaki et al. (Eur. J. Biochem. 269) Ó FEBS 2002 11. Ward, P.N., Field, T.R., Ditcham, W.G., Maguin, E. & Leigh, J.A. (2001) Identification and disruption of two discrete loci encoding hyaluronic acid capsule biosynthesis genes hasA, hasB,andhasC. Streptococcus uberis. Infect. Immun. 69, 392–399. 12. Itano, N., Sawai, T., Yoshida, M., Lenas, P., Yamada, Y., Imagawa, M., Shinomura, T., Hamaguchi, M., Yoshida, Y., Ohnuki, Y., Miyauchi, S., Spicer, A.P., McDonald, J.A. & Kimata, K. (1999) Three isoforms of mammalian hyaluronan synthases have distinct enzymatic properties. J. Biol. Chem. 274, 25085–25092. 13. Tlapak-Simmons, V.L., Kempner, E.S., Baggenstoss, B.A. & Weigel, P.H. (1998) The active Streptococcal hyaluronan syn- thases (HASs) contain a single HAS monomer and multiple car- diolipin molecules. J. Biol. Chem. 273, 26100–26109. 14. Tlapak-Simmons, V.L., Baggenstoss, B.A., Clyne, T. & Weigel, P.H. (1999) Purification and lipid dependence of the recombinant hyaluronan synthases from Streptococcus pyogenes and Strepto- coccus equisimilis. J. Biol. Chem. 274, 4239–4245. 15. Nishida, Y., Knudson, C.B., Nietfeld, J.J., Margulis, A. & Knudson, W. (1999) Antisense inhibition of hyaluronan synthase- 2 in human articular chondrocytes inhibits proteoglycan retention and matrix assembly. J. Biol. Chem. 274, 21893–21899. 16. Camenisch, T.D., Spicer, A.P., Brehm-Gibson, T., Biesterfeldt, J., Augustine, M.L., Calabro, A. Jr, Kubalak, S., Klewer, S.E. & McDonald, J.A. (2000) Disruption of hyaluronan synthase-2 abrogates normal cardiac morphogenesis and hyaluronan- mediated transformation of epithelium to mesenchyme. J. Clin. Invest. 106, 349–360. 17. Nakamura, T., Takagaki, K., Shibata, S., Tanaka, K., Higuchi, T. & Endo, M. (1995) Hyaluronic-acid-deficient extracellular matrix induced by addition of 4-methylumbelliferone to the medium of cultured human skin fibroblasts. Biochem. Biophys. Res. Commun. 208, 470–475. 18. Nakamura, T., Funahashi, M., Takagaki, K., Munakata, H., Tanaka, K., Saito, Y. & Endo, M. (1997) Hyaluronic-acid-defi- cient extracellular matrix induced by addition of 4-methyl- umbelliferone to the medium of cultured human skin fibroblasts. Biochem. Mol. Biol. Int. 43, 263–268. 19. Goldberg, R.L. & Toole, B.P. (1983) Monensin inhibition of hyaluronate synthesis in rat fibrosarcoma cells. J. Biol. Chem. 258, 7041–7046. 20. Smith, T.J. (1990) Retinoic acid inhibition of hyaluronate syn- thesis in cultured human skin fibroblasts. J. Clin. Endocrinol. Metab. 70, 655–660. 21. Zaharevitz, D.W., Chisena, C.A., Duncan, K.L., August, E.M. & Cysyk, R.L. (1993) Vanadate inhibition of hyaluronic acid synthesis in Swiss 3T3 fibroblasts. Biochem. Mol. Biol. Int. 31, 627– 633. 22. August, E.M., Duncan, K.L., Malinowski, N.M. & Cysyk, R.L. (1993) Inhibition of fibroblast hyaluronic acid production by suramin. Oncol. Res. 5, 415–422. 23. Ueki, N., Taguchi, T., Takahashi, M., Adachi, M., Ohkawa, T., Amuro, Y., Hada, T. & Higashino, K. (2000) Inhibition of hya- luronan synthesis by vesnarinone in cultured human myofibro- blasts. Biochim. Biophys. Acta 1495, 160–167. 24. Endo, Y., Takagaki, K., Takahashi, G., Kakizaki, I., Funahashi, M., Yokoyama, M. & Endo, M. (2000) Formation of hyaluronic acid-knock-down extracellular matrix using 4-methylumbellifer- one. In Progress in Transplantation (Munakata, A., ed.), pp. 1–7. Elsevier Science B.V., Amsterdam, the Netherlands. 25. Sohara, Y., Ishiguro, N., Machida, K., Kurata, H., Thant, A.A., Senga, T., Matsuda, S., Kimata, K., Iwata, H. & Hamaguchi, M. (2001) Hyaluronan activates cell motility of v-Src-transformed cells via Ras-mitogen-activated protein kinase and phosphoinosi- tide 3-kinase-Akt in a tumor-specific manner. Mol. Biol. Cell 12, 1859–1868. 26. Chichibu, K., Matsuura, T., Shichijo, S. & Yokoyama, M.M. (1989) Assay of serum hyaluronic acid in clinical application. Clin. Chim. Acta 181, 317–323. 27. DeAngelis, P.L. & Weigel, P.H. (1994) Immunochemical con- firmation of the primary structure of Streptococcal hyaluronan synthase and synthesis of high molecular weight product by the recombinant enzyme. Biochemistry 33, 9033–9039. 28. van de Rijn, I. & Kessler, R.E. (1980) Growth characteristics of group A Streptococci in a new chemically defined medium. Infect. Immun. 27, 444–448. 29. Tanaka, K., Nakamura, T., Ikeya, H., Higuchi, T., Tanaka, A., Morikawa,A.,Saito,Y.,Takagaki,K.&Endo,M.(1994)Hya- luronate depolymerization activity induced by progesterone in cultured fibroblasts derived from human uterine cervix. FEBS Lett. 347, 95–98. 30. Sugahara, K., Schwartz, N.B. & Dorfman, A. (1979) Biosynthesis of hyaluronic acid by Streptococcus. J. Biol. Chem. 254, 6252– 6261. 31. 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. Biochem. 72, 248–254. 32. Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680–685. 33. Sambrook, J., Fritsch, E.F. & Maniatis, T. (1989) Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring. Harbor Laboratory Press, Cold Spring Harbor, New York. 34. Feinberg, A.P. & Vogelstein, B. (1984) ÔA technique for radio- labeling DNA restriction endonuclease fragments to high specific activityÕ. Addendum. Anal. Biochem. 137, 266–267. 35. Towbin,H.,Staehelin,T.&Gordon,J.(1979)Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Natl. Acad. Sci. U.S.A. 76, 4350–4354. 36. Folch, J., Lees, M. & Stanley, G.H.S. (1957) A simple method for the isolation and purification of total lipides from animal tissues. J. Biol. Chem. 226, 497–509. 37. Crater, D.L. & van de Rijn, I. (1995) Hyaluronic acid synthesis operon (has) expression in group A Streptococci. J. Biol. Chem. 270, 18452–18458. 38. Triscott, M.X. & van de Rijn, I. (1986) Solubilization of hya- luronic acid synthetic activity from Streptococci and its activation with phospholipids. J. Biol. Chem. 261, 6004–6009. 39. Goldfine, H. (1972) Comparative aspects of bacterial lipids. Adv. Microb. Physiol. 8, 1–58. 40. Tlapak-Simmons, V.L., Baggenstoss, B.A., Kumari, K., Heldermon, C. & Weigel, P.H. (1999) Kinetic characterization of the recombinant hyaluronan synthases from Streptococcus pyo- genes and Streptococcus equisimilis. J. Biol. Chem. 274, 4246–4253. 41. Klewes, L. & Prehm, P. (1994) Intracellular signal transduction for serum activation of the hyaluronan synthase in eukaryotic cell lines. J. Cell. Physiol. 160, 539–544. 42. Nickel, V., Prehm, S., Lansing, M., Mausolf, A., Podbielski, A., Deutscher, J. & Prehm, P. (1998) An ectoprotein kinase of group C Streptococci binds hyaluronan and regulates capsule formation. J. Biol. Chem. 273, 23668–23673. 43. Jacobson, A., Brinck, J., Briskin, M.J., Spicer, A.P. & Heldin, P. (2000) Expression of human hyaluronan synthases in response to external stimuli. Biochem. J. 348, 29–35. 44. Spicer, A.P., Augustine, M.L. & McDonald, J.A. (1996) Mole- cular cloning and characterization of a putative mouse hyaluronan synthase. J. Biol. Chem. 271, 23400–23406. 45.Itano,N.&Kimata,K.(1996)Molecularcloningofhuman hyaluronan synthase. Biochem. Biophys. Res. Commun. 222, 816–820. 46. Heldermon, C., DeAngelis, P.L. & Weigel, P.H. (2001) Topo- logical organization of the hyaluronan synthase from Strepto- coccus pyogenes. J. Biol. Chem. 276, 2037–2046. Ó FEBS 2002 Inhibition of HA synthesis in Streptococcus by MU (Eur. J. Biochem. 269) 5075 . Inhibition of hyaluronan synthesis in Streptococcus equi FM100 by 4-methylumbelliferone Ikuko Kakizaki 1 , Keiichi Takagaki 1 , Yasufumi Endo 1 ,. effect on their growth. In this study, the inhibition mechanism of hyaluronan synthesis by MU was examined using Strepto- coccus equi FM100, as a model. When