Eur. J. Biochem. 148,479-484 (1985) 0 FEBS 1985

Structural studies on the acidic polysaccharide of Bacillus cereus AHU 1356 cell walls Naoya KOJIMA, Yoshio ARAKI and Eiji IT0 Department of Chemistry, Faculty of Science, Hokkaido University

(Received December 3, 1984/February 6 , 1985) - EJB 84 1268

Structural studies were carried out on the acidic polymer fraction isolated from lysozyme digests of the N - acetylated cell walls of Bacillus cereus AHU 1356. The acidic polymer fraction contained glucosamine, galactose, rhamnose, glycerol and phosphorus in a molar ratio of 1 : 1 :2: 1 : 1 , together with small amounts of glycopeptide components and muramic acid 6-phosphate. The hydrogen fluoride treatment led to removal of glycerol and phosphorus from the polymer without loss of other components. Results of the NaIO, oxidation, methylation and proton magnetic resonance spectroscopy of the native and dephosphorylated preparations, in combination with data of the analysis of oligosaccharides obtained from partial hydrolysis of the polysaccharide, led to the most likely structure of the repeating units of the acidic polysaccharide chain, --f 4)N-acetylglucosaminyl- (LY 1 --f 3)rhamnosyl(a 1 -+ 3)galactosyl(cr 1 -+ 4)[sn-glycerol l-phospho-2]rhamnosyl(a 1 4.

In a previous paper [l], we reported the separation of a major, neutral polysaccharide fraction and a minor, acidic polysaccharide fraction from lysozyme digests of Bacillus cereus AHU 1356 cell walls. The neutral polysaccharide was shown to have a complicated structure composed of N- acetylglucosamine, N-acetylmannosamine, N-acetylgalactos- amine and glucose. On the basis of preliminary analytical data, the acidic polysaccharide appeared to be a kind of glyc- erol teichoic acid. However, either the NaIO, oxidation or hydrogen fluoride treatment of this polysaccharide led to complete loss of glycerol residues without degradation of the polymer chain, suggesting that this polysaccharide is a unique polymer with glycerol phosphate residues attached as side branches. This fact prompted us to elucidate the structure of the acidic polysaccharide. The present paper describes results of the structural studies on the acidic polysaccharide components of B. cereus AHU 1356 cell walls.

MATERIALS AND METHODS

Acidic-polysaccharide-linked glycopeptide

Acidic-polysaccharide-linked glycopeptide (denoted as AP-GP) was prepared from lysozyme digests of the N- acetylated cell walls of Bacillus cereus AHU 1356 by the same procedures as those in a previous paper [l].

Correspondence to E . Ito, Department of Chemistry, Faculty of Science, Hokkaido University, Kita-I 0-jyo, Nishi-8-chome, Kita-ku, Sapporo, Japan 060

Abbreviations. AP-GP, acidic-polysaccharide-linked glycopep- tide; Rha, rhamnose

Enzymes. Glycerol-3-phosphate dehydrogenase (EC 1.1 .I 3); alkaline phosphatase (EC 3.1.3.1); acid phosphatase (EC 3.1.3.2); lysozyme (EC 3.2.1.17); a-L-rhamnosidase (EC 3.2.1.40); cc-D-galac- tosidase (EC 3.2.1.22); a-N-acetyl-D-glucosaminidase (EC 3.2.1.50).

Hydrolysis of acidic-polysaccharide-linked glycopeptide with hydrogen,fluoride

The AP-GP preparation (1 mg) was treated in 0.2 ml47% hydrogen fluoride (HF) at 4°C for 30 h, 72 h or 120 h. After evaporation to dryness by an air flash, the product was dis- solved in 0.5ml 50mM NH4HC03 and subjected to chromatography on a Sephadex G-25 column (1 x 80 cm) in the same salt solution. Fractions (0.6 ml) were collected and analyzed for hexose, phosphorus and free glycerol.

To prepare a dephosphorylated polysaccharide prepara- tion, AP-GP (30 mg) was treated in 3 ml 47% HF at 4°C for 40 h and then chromatographed as above. The polymer fraction, which was excluded from the column, retained the majority of hexose, but contained only negligible amounts of glycerol and phosphorus. This polymer fraction was used as the dephosphorylated polysaccharide preparation (25 mg) after lyophilization.

Periodate oxidation

The AP-GP or dephosphorylated polysaccharide prepara- tion (1 mg each) was oxidized with 35 mM NaIO, in 1 ml sodium acetate buffer (pH 4.5) for 3 days in the dark at 4°C. After destruction of residual NaIO, by the addition of ethyl- ene glycol, the product was reduced by treating with NaBH, at room temperature overnight and then dialyzed. The resulting nondialyzable fraction was hydrolyzed at 100°C either in 1 M HCl for 2 h (for analysis of neutral sugars) or in 4 M HC1 for 4 h (for analysis of amino sugars).

Partial acid hydrolysis

The AP-GP preparation (4 mg) was treated in 1 mlO.1 M HCl at 100°C for 15 or 60 min. After lyophilization, the product was applied on a DEAE-cellulose column (1.5 x 3 cm) and the column was eluted with water, then with a linear

480

gradient of NH4HC03. Fractions containing both hexose and phosphorus were pooled and further fractionated into three phosphorus-containing saccharides by chromatography on a Sephadex G-25 column (1 x 80 cm).

Methylation analysis

The dephosphorylated polysaccharide preparation (10 mg) was permethylated by the method of Hakomori [2] and hydrolyzed in 5% H2S04 at 100°C for 4 h . The hydrolysate was passed through a Dowex 1 column (acetate form) and applied on a Dowex 50 column (H' form). The column was eluted with water (methylated neutral sugar frac- tion), then with 1 M HC1 (methylated amino sugar fraction). The methylated neutral sugar was analyzed as its alditol ace- tate derivative on ECNSS-M in a glass column at 180°C [3], and the methylated amino sugar was analyzed on 5% SE-52/ Chromosorb W in a glass column after N-acetylation and trimethylsilylation [4]. 3,6-Di-O-methyl-N-acetylglucosamine and 4,6di-O-methyl-N-acetylglucosamine were prepared from chitobiose and hyalobiose [5], respectively, and used as standards.

Other materials and analytical methods

Unless otherwise indicated, materials and methods were the same as those described in previous papers [l, 6, 71. sn- Glycerol-3-phosphate dehydrogenase was obtained from Sigma Chemical Co. ; exo-a-galactosidase from Seikagaku Kogyo Corp. Exo-a-N-acetylglucosaminidase was prepared from human urine as described by Figura [8]. Amino acids, amino sugars and muramic acid 6-phosphate were analyzed on a Shibata AA-autoanalyzer after acid hydrolysis [I]. Total hexose was analyzed by the phenol/H2S04 method with glucose as a standard [9]; phosphorus, by the method of Lowry et al. [lo]; formaldehyde, with chromotropic acid [ll]; glycerol, by the method of Wieland [12]. Glycerol, hexose, polyol and hexosamine were analyzed by gas-liquid chromatography after N-acetylation and trimethylsilylation [6, 71. Paper chromatography was carried out by the des- cending method on Toyo no. 50 filter paper in 1-butan011 pyridine/water (6: 4: 3, v/v/v). Paper electrophoresis was performed on the same paper in pyridine/acetic acid/water (35: 5:960, v/v/v, pH 5.8) at 50 V/cm for 2 h. The measure- ment of the proton magnetic resonance spectrum was carried out with a Jeol FX-400 spectrometer at 25 "C or 95 "C in a 3.5% solution. Prior to analysis, samples were repeatedly treated with D20 by dissolving in it and lyophilization. Chemical shifts are given relative to an internal standard, sodium-3-methylsilylpropane sulfonate.

RESULTS Composition of acidic-polysaccharide-linked glycopeptide

The neutral-polysaccharide-linked and acidic-polysac- charide-linked glycopeptides were separated from lysozyme digests of the N-acetylated cell walls of B. cereus AHU 1356 by ion-exchange chromatography and gel chromatography, as reported previously [l]. The yields of the former and the latter complexes, respectively, were 360 mg/g and 90 mg/g cell walls. The acidic-polysaccharide-linked glycopeptide prepara- tion (AP-GP) was shown to contain galactose, excess glucosamine over the amount of muramic acid derivatives, rhamnose, glycerol and phosphorus in amounts of 925, 1080,

I ! 1 vo

0.4 - L L

3 I

2 1

I

\ 5 0.2

0.0 5 -

0.41

6

40 50 60 70 Fraction number

Fig. 1. Hydrolysis of acidic-polysaccharide-linked glycopeptide with hydrogenfluoride. After the hydrolysis of AP-GP (1 mg) with 47% HF at 4°C for 30 h (A), 72 h (B) and 120 h (C), each product was chromatographed on Sephadex G-25. (0) Hexose; (0 ) phosphorus; ( A ) free glycerol. Arrows 1 , 2 and 3 indicate the elution positions of standards, glucose, maltotriose and maltopentaose, respectively

1880, 900 and 980 nmol . mg-', respectively, together with small amounts of glycopeptide components, as analyzed mainly by an autoanalyzer and gas-liquid chromatography. These components account for 96% of the weight of the preparation. The lower value (86%) reported in the previous paper [l] may be explained by missing the rhamnose residues. The apparent molecular mass of AP-GP was about 100 kDa, as estimated by gel chromatography on Sepharose 6B.

Dephosphorylation and partial degradation of polysaccharide by hydrogen fluoride treatment

Fig. 1 shows the process of partial degradation of AP-GP by HF hydrolysis. Until 30 h, glycerol and phosphorus were released from the polysaccharide without appreciable cleavage of the glycosidic linkages in the polymer chain, The average molecular mass of the polymer was somewhat reduced after the treatment for 40 h, though the value still exceeded 20 kDa as estimated by gel chromatography on Sephacryl S-200. After 72 h, the amount of polymer excluded from the column decreased markedly, and a tetrasaccharide and a higher oligosaccharide appeared (Fig. 1 B). The treatment for 120 h led to the maximum production of the tetrasaccharide with the formation of small saccharides and nearly complete disappearance of the polymer (Fig. 1C). These results show that the HF treatment of AP-GP for short periods caused specific cleavage of phosphodiester bonds, whereas the pro- longed HF treatment led to rather nonspecific cleavage of acid-labile glycosidic bonds involved in the backbone chain.

48 1

Rha H-l

l . . . . l 1 . . . I . . . . I . . . . I . . . . I . . . . l , 6.0 5.0 4.0 3.0 2.0 1.0 0.0

PPm

Fig.2. Proton N M R spectrum taken at 95°C with a 400-MHz instru- ment

Thus, the dephosphorylated polysaccharide was prepared by the HF treatment of AP-GP for 40 h. This preparation contained negligible amounts of glycerol, phosphorus and glycopeptide components.

Proton magnetic resonance spectroscopy

The anomeric configuration of the saccharide constituents of the AP-GP and dephosphorylated polysaccharide prepara- tions was studied by proton magnetic resonance spectroscopy in deuterium oxide solutions. The spectrum given by AP-GP exhibited signals corresponding to four different anomeric protons (6 = 5.100, 5.070, 5.017 and 4.931 ppm), together with signals for the acetamide groups of N-acetylglucosamine residues (6 = 2.052 ppm, 3H) and for the methyl groups of rhamnose residues (6 = 1.347 and 1.302 ppm, 3H each) (Fig.2). The doublet signals at 6 = 5.100 and 5.017 ppm are assigned to a-glycosidically linked N-acetylglucosmine and galactose residues, respectively, and this assignment is in accordance with the coupling constants (J = 3.29 Hz and 3.62 Hz). The semidoublet signals at 6 = 5.070 ppm and 4.931 ppm are tentatively assigned to a-glycosidically linked rhamnose residues on the basis of their chemical shifts and the spectral data of methyl a-rhamnoside [13]. The signal of rhamnose residues in a relatively high field region (6 = 4.931 ppm) may be explained by a shielding effect of phosphoryl substitution on these residues, because this signal shifted to a downfield region (6 = 4.952 ppm) in the spectrum of the dephosphorylated polysaccharide preparation.

Owing to the manno-configuration of rhamnose, chro- mium trioxide oxidation gave no reliable information con- cerning anomeric configuation of this saccharide [14]. On the other hand, the absence of signals from higher-field regions between 6 = 3.30 ppm and 3.50 ppm was consistent with the assignment of a-rhamnosyl residues, since the signal of the H-5 proton of P-rhamnoside was reported to be present in these high-field regions by Bruyn et al. [15]. Thus, all of the saccharide constituents of the polysaccharide are believed to be linked a-glycosidically.

Methylation analysis

The dephosphorylated polysaccharide preparation was hydrolyzed after permethylation. The hydrolysate was

fractionated into two fractions, methylated neutral sugar and methylated amino sugar fractions, by chromatography on a column of Dowex 50 (H' form). When the methylated neutral sugar fraction was analyzed by gas-liquid chromatography after conversion into alditol acetates, this fraction was shown to contain 2,4,6-tri-U-methylgalactitol acetate and 2,3-di-U- methylrhamnitol acetate and/or 2,4-di-U-methylrhamnitol acetate, from comparison of the retention times with those reported [3]. The molar ratio of the galactitol derivative and the rhamnitol derivative(s) was about 1 : 2. The component of the methylated amino sugar fraction was coincident with standard 3,6-di-U-methyl-N-acetylglucosamine, but not with 4,6di-U-methyl-N-acetylglucosamine. The amount of 3,6di- U-methyl-N-acetylglucosamine accounted for about 70% that of the glucosamine residues present in the polysaccharide. Thus, the polymer seems to consist of 4-substituted glucosamine, 3-substituted galactose and 3-substituted and/ or 4-substituted rhamnose residues.

Periodate oxidation of acidic polysaccharide before and after dephosphorylation

The NaI04 oxidation of AP-GP led to the complete degra- dation of the glycerol residues with the formation of an equimolar amount of formaldehyde. The oxidation product gave the same composition as did AP-GP except for the ab- sence of glycerol, as analyzed by gas-liquid chromatography after acid hydrolysis followed by N-acetylation. The NaI04 oxidation of the dephosphorylated preparation gave a nondialyzable product which contained equimolar amounts of galactose, glucosamine and rhamnose, indicating that half of the rhamnosyl residues were converted to an NaI04- sensitive form by dephosphorylation. These results suggest that the polysaccharide chain in AP-GP was composed of equimolar amounts of galactose, glucosamine, nonphos- phorylated rhamnose and phosphorylated rhamnose. The nonphosphorylated rhamnose and phosphorylated rhamnose residues, respectively, seemed to correspond to the 3-sub- stituted and 4-substituted rhamnosyl residues demonstrated by methylation analysis.

Isolation and characterization of oligosaccharides

The products resulting from hydrolysis of AP-GP in 1 ml 47% H F at 4°C for 5 days were fractionated by chro- matography on Sephadex G-25 with an elution profile similar to that shown in Fig. 1C. Fractions corresponding to the bars indicated in Fig. 1C were pooled and further purified by re- chromatography on the same column and subsequent paper chromatography. Yields of sacharides 2, 3, 4, 5 and 6 were 1.5, 0.65, 10.6, 6.7 and 3.0 Fmol as hexose, respectively.

The composition of these saccharides are summarized in Table 1. Saccharide 1 was a monosaccharide mixture com- posed mainly of rhamnose and galactose. Saccharides 2 and 3 were shown to be disaccharides, Gal-Rha and GlcNAc-Rha, respectively. Saccharide 2 was sensitive to digestion with or- galactosidase, whereas saccharide 3 was sensitive to a-N- acetylglucosaminidase. Both saccharides gave equimolar amounts of formaldehyde by the NaI04 oxidation after reduc- tion with NaBH4. These results indicate that the rhamnose residue in either saccharide was unsubstituted at C-2. Thus, the tentative structures for saccharides 2 and 3 are Gal(a 1 + 3/ 4)Rha and GlcNAc(a 1 --f 3/4)Rha, respectively.

Saccharide 4 was a linear tetrasaccharide composed of galactose, glucosamine and rhamnose in a molar ratio of

482

Table 1. Composition of oligosaccharides Saccharides 2-6 were obtained from H F hydrolysate of AP-GP and analyzed after acid hydrolysis. Components were also analyzed after reduction of the saccharides with NaBH4 (reduction). Reduced saccharides 4 and 5 were subjected to oxidation with NaI04 followed by reduction with NaBH4, and the products were analyzed (oxidation). Data are shown in molar ratios to galactose except for the case of saccharide 3, in which data are shown in molar ratios to glucosamine

Saccharide Treatment Molar ratio

galactose rhamnose glucosamine glycerol rhamnitol

2 None

3 None

4 None

Reduction

Reduction

Reduction Oxidation

Reduction Oxidation

Reduction

5 None

6 None

1 .o 1 .o 0 0 1 .o 1 .o 1 .o 1 .o 1 .o 1 .o 1 .o 1 .o

0.9 0.1 0.9 0 2.1 1.1 1.1

2.1 1.5 1.1 1.9 1.7

0 0

1 .o 1 .o 1.1 0.9 0.1 1.1 0.9 0.5 0.9 0.9

0 0

0 0 0 0 0.8

0 0 0.4 0 0

0 1 .o 0 0.9

0 0.9 0 0 0.5 0

0 0.2

1:1:2 with a rhamnose residue at the reducing end. The NaIO, oxidation of this saccharide after reduction with NaBH4 led to degradation of the glucosamine residue, leaving the galactose residue and one rhamnose residue undegraded. Therefore, it seems that the latter two residues were sub- stituted at C-3 and present in inner positions of the tetrasaccharide. The formation of one equivalent formal- dehyde was also observed and is accounted for by the oxida- tion of the reducing terminal rhamnose. Furthermore, when saccharide 4 was treated with 47% HF at 4°C for 5 days, it gave two disaccharides; one was identical with disaccharide 2, and the other with saccharide 3. These results lead to the most likely structure of saccharide 4, GlcNAc(cr 1 4 3)- Rha(a 1 -+ 3)Gal(cc 1 --f 3/4)Rha.

Saccharide 5 was a linear octasaccharide with a glucos- amine residue and a rhamnose residue at the nonreducing and reducing ends, respectively, and was composed of galactose, glucosamine and rhamnose in a molar ratio of 1 : 1 : 2, the same as that in saccharide 4. HF hydrolysis (4"C, 5 days) of saccharide 5 gave a tetrasaccharide as the major product and two disaccharides as minor products. The major product was shown to be identical with saccharide 4, while the minor dis- accharides to be saccharides 2 and 3. Thus, saccharide 5 was most likely a dimer of saccharide 4. The NaI04 oxidation of this saccharide led to degradation of half each of the rhamnose and glucosamine residues. The NaI0,-sensitive and the NaI0,-resistant glucosamine residues seem to be present at the nonreducing end and at the inner position of the saccharide, respectively. The NaIO,-resistant rhamnose re- sidues (2 mol) seem to be substituted at C-3, and the inner NaI04-sensitive one seems to be substituted at C-2 or C-4. Taking account of nonsubstitution at C-2 of the NaI04- sensitive, reducing terminal rhamnose residues in saccharides 2 and 4, the inner NaI04-sensitive rhamnose as well as the reducing terminal one in saccharide 5 seems to be substituted at C-4. Thus the most likely structure for saccharide 5 is GlcNAc(cr 1 --f 3)Rha(cc 1 -+ 3)Gal(cc 1 --* 4)Rha(cc 1 + 4)GlcNAc- (a 1 -+ 3)Rha(cr 1 -+ 3)Gal(cc 1 -+ 4)Rha. Saccharide 6 showed the same composition as saccharide 4 or 5, but further structural studies on this saccharide were not performed.

I .6

I .2

0.8

3 0.4 L u..

0

0

\

f - "'0 10 20 30 40 50

2 Fraction number

- r 0

- rrl

.3 y P I

.o

Fraction number

Fig. 3. Isolation of phosphorylated saccharide fragment. (A) AP-GP (4mg) was hydrolyzed in 0.1 M HCl at 100°C for 15 min. After lyophilization, the product was chromatographed on a DEAE- cellulose column (1.5 x 3 cm). The column was eluted with water, then with a linear gradient of NH4HC03 (0-0.3 M). Fractions (1 ml) were collected and analyzed for hexose (0) and phosphorus (0) . (B) The pooled fraction indicated by bar in upper panel was chromatographed under the conditions same as described in Fig. 1. Symbols and arrows are the same as in Fig. 1

The data described above suggest that saccharide 4 was the repeating units of the backbone chain of dephosphorylated polysaccharide. Taking account of the above described results of the NaIO, oxidation, the position of phosphoryl substitu- tion in AP-GP seems to be either C-2 or C-3 of the 4-sub- stituted rhamnose residue.

483

Table 2. Composition of phosphorylated saccharide,fragment A phosphorus-containing saccharide fragment was obtained from the partial acid hydrolysate of AP-GP (0.1 M HCI, IOO'C, 15 min) by the same procedure as described in Fig.2B. After purification by paper electrophoresis, the fragment was analyzed before and after reduction with NaBH4. HCHO formation in NaIO, oxidation was also determined. Data are shown in molar ratios to phosphorus

Component Molar ratio

before after reduction reduction

Phosphorus 1 .oo 1 .oo Rhamnose 0.97 0 Rhamnitol 0 0.95 Glycerol 1.01 1.03 HCHO formed by NaIO, oxidation 0.94 0.95

Table 3. Identification of sn-glycerol 1-phosphate The glycerol phosphate fraction, obtained from the partial acid hydrolysate of AP-GP (0.1 M HCI, 100°C, 60 min), was analyzed. Data are shown in molar ratios to the phosphorus content in the sample

Analysis Molar ratio

Phosphate liberation by digestion with alkaline phosphatase 0.96

HCHO formation in NaI04 oxidation (corresponding to the amount of cc-glycerol phosphate) 0.96

Formation of oxidized product by treatment with sn-glycerol-3- phosphate deyhdrogenase 0.14

Isolation and characterization ofphosphorylated saccharide fragment

The acid hydrolysate (0.1 M HCI, 1OO"C, 15 min) of AP- GP was subjected to ion-exchange chromatography on a DEAE-cellulose column (Fig. 3A), and the resulting second fraction, which contained both phosphorus and hexose, was further fractionated by chromatography through a Sephadex G-25 column (Fig. 3B). Preparative paper electrophoresis of the component in the third peak gave a phosphorylated saccharide (0.98 pmol as phosphorus). This fragment, which was composed of rhamnose, glycerol and phosphorus in an equimolar ratio (Table2), was identified to be glycero- phospho-2-rhamnose from the following evidence. It gave an equimolar amount of rhamnitol as analyzed by gas-liquid chromatography after reduction with NaBH,. The NaIO, oxidation of this fragment carried out either before, or after reduction with NaBH, gave an equimolar amount of formaldehyde.

Identification of glycerol phosphate

When the AP-GP preparation was treated with 0.5 M NaOH at 100°C for 80 min, the glycerol residues were quantitatively liberated from the polymer, but glycerol phosphate was given only in a negligible amount. The polymer fraction resulting from the alkaline treatment showed the same composition as did the untreated sample, except for the absence of glycerol.

We tried to isolate glycerol phosphate from the partial acid hydrolysate of AP-GP, though migration of phosphoryl groups is known to occur under these acidic conditions. The AP-GP preparation (4 mg) was hydrolyzed in 0.5 ml 0.1 M HCl at 100°C for 1 h, then glycerol phosphate was isolated from the hydrolysate by successive chromatography on columns of DEAE-cellulose and Sephadex G-25. The yield of glycerol phosphate was 0.45 pmol. Analysis involving phosphatase digestion, NaIO, oxidation and dehydrogena- tion with sn-glycerol-3-phosphate dehydrogenase, are sum- marized in Table3. The result suggests that the majority of the glycerol phosphate residues were sn-glycerol 1 -phosphate.

DISCUSSION

The above results lead to the most likely structure of the repeating units of the acidic polysaccharide of Bacillus cereus AHU 1356 cell walls, --t 4)GlcNAc(a 1 + 3)Rha(a 1 + 3)Gal- (a 1 + 4) [sn-glycerol-l-phospho-2]Rha(a 1 +. The alkaline treatment led to liberation of glycerol residues from the acidic polymer chain without loss of phosphorus, indicating that the cleavage of the linkages between the glycerol residues and the phosphoryl groups occurred through 2,3-cyclic phosphodi- ester formation on the rhamnose residues. This result is in agreement with the presumption that the phosphorylated rhamnose residues were 4-substituted, but not 3-substituted.

The proton magnetic resonance spectroscopic data indi- cate that all of the saccharide consitutents are linked a- glycosidically. The assignment of a-rhamnosyl residues was also supported by the finding that the saccharide moieties resulting from mild acid hydrolysis of Rha-Gal-linked and Rha-GlcNAc-Rha-Gal-linked lipids which are involved in the synthesis of the acidic polysaccharide of this organism were sensitive to digestion with a-rhamnosidase (unpublished ob- servation).

The acidic polysaccharide of this organism is a novel type of wall polysaccharide with sn-glycerol 1 -phosphate residues linked as the side branches of the polymer chain. Such periph- eral location of the glycerol phosphate residues has been re- ported in the membrane-derived oligosaccharides [16] and the polysaccharide surface antigens [17] of Escherichia coli and the acidic polysaccharide of Streptococcus pneumoniae [ 181. Recently, sn-glycerol 1-phosphate has been revealed to be a component of the N-acetylmannosaminuronic-acid-contain- ing polysaccharide of Bacillus subtilis AHU 1031 [19]. The glycerol phosphate residues in the polysaccharide surface anti- gens of E. coli were shown to be an immunodeterminant [17].

The sn-glycerol 1-phosphate residues in the lipoteichoic acids of gram-positive bacteria seem to be derived from the glycerol phosphate moiety of phospholipids, such as phosphatidylglycerol. In addition, biosynthetic studies on the membrane-derived oligosaccharides of E. coli showed that the sn-glycerol 1 -phosphate residues in these materials arise from the glycerol phosphate moiety of phosphatidylglycerol [20, 211. Our preliminary pulse-chase experiment indicated that the sn-glycerol 1 -phosphate residues of the acidic polysaccharide of B. cereus AHU 1356 cell walls also arise from phosphatidylglycerol (unpublished data). Further bio- synthetic studies on the acidic polysaccharide of this organism are in progress and will be described elsewhere.

REFERENCES 1. Amano, K., Hazama, S., Araki, Y. & Ito, E. (1977) Eur. J .

Biochem. 75, 51 3 - 522.

2. Hakomori, S. (1964) J . Biochem. (Tokyo) 55,205-208. 3. Lindberg, B. (1972) Methods Enzymol. 28, 178-195. 4. Sweeley, S. S., Bentley, R., Makita, M. &Wells, W. W. (1963) J .

5. Weissman, B. & Meyer, K. (1954) J . Am. Chem. SOC. 76, 1753-

6. Sasaki, Y., Araki, Y. & Ito, E. (1983) Eur. J . Biochem. 132,207-

7. Kojima, N., Araki, Y & Ito, E. (1983) J . Biol. Chem. 258,9043 -

8. Figura, K. (1977) Eur. J . Biochem. 80, 525-533. 9. Dubois, M., Gilles, J. K., Hamilton, J. K., Roberts, P. A. &

Smith, F. (1956) Anal. Chem. 28, 350 - 356. 10. Lowry, 0. H., Roberts, N. R., Leiner, K. W., Wu, M.-L. & Farr,

L. (1954) J . Biol. Chem. 207, 1 - 17. 11. Hanahan, D. J. & Olley, J. N. (1958) J . Biol. Chem. 231, 813-

828. 12. Wieland, 0. (1974) in Methods in enzymatic analysis (Bergmeyer,

H. V., ed.) 2ndedn, pp. 1401 - 1404, Verlag Chemie, Weinheim/ Bergst.

Am. Chem. SOC. 85,2497 - 2507.

1757.

213.

9045.

13. Jansson, P. E., Kenne, L., Liedgren, H., Lindberg, B. & Lonngren, J. (1976) Chem. Commun. Univ. Stockholm 8, 46- 120.

14. Lindberg, B. & Lonngren, J. (1979) Methods Enzymol. 50C, 3- 33.

15. Bruyn, A. D., Antennis, M., Gussem, R. D. & Dutton, G. G. S. (1976) Carbohydr. Res. 47, 158-163.

16. Schneider, J. E., Reinhold, V., Rumley, M. K. & Kennedy, E. P. (1979) J. Biol. Chem. 254, 10135-10138.

17. Jann, B., J a m , K., Schmidt, G., Oerskov, I. &Oerskov, F. (1975) Eur. J . Biochem. 15,29-39.

18. Estrada-Parra, S. & Heiderberger, M. (1963) Biochemistry 2,

19. Yoneyama, T., Araki, Y. & Ito, E. (1984) Eur. J . Biochem. 141,

20. Goldberg, D. E., Rumley, M. K. & Kennedy, E. P. (1981) Proc.

21. Jackson, B. J . &Kennedy, E. P. (1983) J . Biol. Chem. 258,2394-

1288 - 1294.

83 - 89.

Natl Acad. Sci. USA 78, 5513-5517.

2398.