N-unsubstituted glucosamine residues and other modifications in murein of the obligatory chemolithotroph Thiobacillus neapolitanus

© INSTITUTP.~,STEUR/ELSEVIER Paris 1992

Res. MicrobioL 1992, 143, 47-54

N-unsubsfituted glucosamine residues and other modifications in murein of the obligatory chemolithotroph Thiobacillus neapolitanus J. Baj, I. Grabowska and Z. Markiewicz (*) Institute of Microbiology, University of Warsaw, Nowy Swiat 67, 00-046 Warsaw 64 SUMMARY Purified murein from Thiobacillus neapolitanus was poorly digested by lysozyme. It's sensitivity to the enzyme greatly increased after N-acetylation. The murein was found to contain 30 to 3 5 % glucosamine residues lacking N-acetyl groups. It also contained phosphomuramic acid. Further modifications included amidation of diaminopimelic acid in the peptide side chains and a low alanine content. None of these modifications were found in the murein of another sulphur bacterium, Thiobacillus versutus.

Key-words: Thiobacillus neapofitanus, Peptidoglycan, Glucosamine, Lysozyme; DAP, N-acetyl, Chemolithotrophy.

INTRODUCTION The murein sacculus is a unique polymer in terms of structure, formation and function. It serves to maintain the integrity of the protoplast, at the same time determining the shape of the cell. The mureins (peptidoglycans) of individual species may differ in structure. The differences can involve the composition of both the glycan strands and the peptide moiety, the degree of cross-linking and the thickness of the murein layer. The primary chemical composition of the glycan chains is highly conserved, though the following modifications have been described: (1) acetylation or phosphorylation of 6-hydroxymuramic acid groups, (2) lack of acetylation of the amino groups on C2 of glucosamine (GlcN) or muramic acid (Mur), (3) substitution of the Submitted March 15, 1991, accepted July 12, 1991. (*) Correspondingauthor.

N-acetyl group in Mur by N-glycolyl, (4) internal cyclization of N-acetylmuramic acid (MurNac) residues at chain termini. The lack of amino sugar N-acetylation and the presence of O-phosphorylation had been claimed to be typical of only Gram-positive bacteria. However, in recent years similar modifications have been described for certain Gram-negative species (Schmelzer et aL, 1982; Jiirgens et aL, 1987b, Woitzik et aL, 1988). The peptide moiety of murein shows considerable variation and almost 100 different primary structures have been found (Schleifer and Kandler, 1972; Schleifer and Stackebrandt, 1983; Schleifer and Seidl, 1985). The greatest variation is found in position 3 of the stem peptide which is usually taken up by a diamino acid (e.g. diaminopimelic acid (DAP) or lysine) but also

48

J. 8 A J E T A L .

by ornithine or meso-lanthionine (Rogers et al., 1980). Variation in the peptide side chains and bridges is particularly frequent in Gram-positive bacteria, which has allowed the murein of this group to be regarded as an important chemotaxonomic criterion (Schleifer a n d Seidl, 1985). Moreins of most Gram-negative eubacteria have been classified as type A l ' r (Schleifer and Kandler, 1972). They contain D A P (in position 3) which forms a direct cross-bridge with Dalanine o f a second side chain. In the last decade, mureins of several photosynthetic bacteria have been characterized. Some of them seem to be particularly interesting because of their modifications which until now have been claimed to be typical o f only Gram-positive bacteria. Very little is k n o w n a b o u t the peptidoglycan of the second group of autotrophic bacteria, the chemolithotrophs. In recent papers, we described the murein structure of Thiobacillus versutus, a facultative chemolithotrophic species (Baj and Markiewicz, 1988) and presented some preliminary results concerphng the murein of Thiobacillus neapolitanus, an obligatory chemolithotroph (Baj and Markiewicz, 1991). The murein of the latter was found to contain glutamic acid (Glu), alanine (Ata), D A P , Mur and GlcN, constituents typical of most mureins from other Gram-negative species. A m i n o acid and amino sugar analysis also revealed the presence of two c o m p o u n d s with retention times indicative of modified amino sugars, as well as a high a m o u n t o f a m m o nia. The identification of these compounds is the subject of the present paper.

from Dr. Tuovinen of the University of Groningen. T. neapolitanus was grown with intensive shaking at 28°C in mineral medium containing: 0.5070 Na2S203.5H20, 0.15070 KH2PO 4, 1.59°70 Na2HPO 4, 0.04070 NH4CI, 0.01o70 MgSOn.7HzO pH7.5, supplemented with microelements (Vishniac and Santer, 1957). Cells were harvested at A57s = 0.2. T. versutus was grown with vigorous shaking at 30°C. Escherichia coli KNI26 was from Dr. Hfltje of the MPI, Tiibingen. Bacillus cereus AHUI356 was kindly provided by Dr. Ito of Hokkaido University. Both strains were grown with shaking at 37°C.

Preparation of murein

Isolation of murein from E. coli, T. versutus and T. neapolitanus followed the procedure described by Glauner (1988). For some experiments, murein from T. versutus was additionally treated with 2.5 M NaOH to remove contaminating protein (Poindexter and Hagenzieker, 1981). Murein from B. cereus was a kind gift of Dr. E. Ito. For some experiments it was isolated by us as previously described (Araki et al., 1972). Purified mureins were kept at 2°C in bidistilled water containing 0.02°70 sodium azide.

Dinitrophenylation of murein and amino sugars

Dinitrophenylation of murein v/as essentially as described by Schmelzer et al. (1982). Standards for thin layer chromatography of acid hydrolysates of dinitrophenylated murein were prepared by treating 2 mg of Mur or GlcN in 0.5o70 NaHCO 3 with 25 mg l-fluoro-2,4-dinitrobenzene in 1 ml ethanol. After overnight incubation in the dark, the mixture was extracted with ether and the aqueous phase was lyophilized, resuspended in distilled water and stored frozen (Schleifer, 1985, modified). N-acetylation of mv~rein

MATERIALS AND METHODS

N-acetylation ,~vas performed according to Hayashi et al. (19'73).

Bacterial strains and growth conditions Colorimetrie determination of O-acetyl groups

All cultures, except for T. neapolitanus were grown to mid-log phase in Luria broth (Miller, 1972). T. neapolitanus and T. versutus strai~: were obtained

O-acetyl grc, ups in murein were assayed according to the method of Hestrin (1949). A 4 mM solu-

DAP DNP BTI GIcN

GlcNac = Mur = MurNac = Mur-P =

= = = =

diaminopimelic(acid) dinitrophenyl. (I.Dtrifluoroacetoxy)iodobenzene. glucosamine.

glucosamineN-acetylated. rnuramicacid. N-acetylmuramicacid. Mar-phosphate.

MUREIN MODIFICATIONS IN THIOBACILLUS NEAPOLITANUS tion of acetylcholine chloride in I mM sodium acetate pH 5.5 was used as a standard.

Determination of acetamide groups The presence of acetamide groups in peptide side chains of murein was identified according to Soby and Johnson (1981). Essentially, murein in 10 mM CF3COOH was supplemented with an equal volume of bis (l,l-trifluoroacetoxy)iodo-benzene (BTI) solution in dimethylformamide (36 mg/mi). The samples were sealed and incubated for 4 h in the dark at 60°C. Murein was centrifuged, extracted with n-butyl acetate to remove residual BTI and suspended in 4 M HCI for acid hydrolysis (105°C, 16 h). The products of hydrolysis were separated as above.

49

Quantitation of muramic acid Mur content was estimated by the method of Hadzija (1974) which makes use of the degradation of Mur to lactic acid and then acetic antlydride. The latter gives a coloured reaction with p-hydroxydiphenyl. Absorbance was measured at 560 nm. Calculations were made on the assumption that the weight content of lactic acid in muramic acid was 30070. Dcterminatioa of hexo~amine The colorimetric method of Tsuji et al. (1969) with a sensitivity of 1 p.g/ml was used. The results were referred to a standard curve for GleN.

RESULTS Lysozyme digestion of murein Murein was suspended in 20 mM Tris-HCl buffer pH 7.4. Lysozyme was added to obtain an enzyme to murein ratio of not more than 1/50 (w/w). The mixture was shaken at 37°C. Degradation of murein was estimated by measuring the absorbance of the preparation at 500 nm and following the increase in the amount of reducing sugars.

Determination of reducing groups Reducing sugars liberated by the action of lysozyme were quantitated as described by Thompson and Shockman (1968). A standard curve was prepared using GlcNac.

Isolation of muramic acid phosphate (Mur-P) An acid hydrolysate of murein (4 h, t00°C, 4 M HCI) was, after careful removal of HCI, dissolved in water. Insoluble residue was removed by centrifugation and the ~t!perr~a!~nt was concentrated. The hydrolysate was applied to a "Dowex 50WX8" column, H + form (0.9 × 12 cm), prepared according to the manufacturer (Serva, FRG). The material loaded on the column was eluted first with redistilled water, then with 0.4 M HCI. Fractions of 2 ml were collected.

Determination of phosphate The method used was that of Lowry et aL (1954). All glassware was washed in chromic acid and rinsed with glass-bidistilled water.

Amino acid and amino sugar composition of murein Extraction o f T. neapolitanus with boiling 4% sodium dodecyl sulphate resulted in the isolation o f a practically protein-free murein fraction as described in detail elsewhere (Baj et al., 1991). A m i n o acid and amino sugar analysis showed the molecular ratio o f the murein components to be approximately GIcN:Mur:DAP: Ala:Glu = 0.4:0.55:0.94:t. 19:1.0. However, the acid hydrolysate yielded two unidentified components (fig. 1). One was eluted from the column before Mur (X), the second just after GleN (Y). Another finding was the high amount of ammonia (between 1 and 1.75 mol per mol Glu, depending on the experiment). The latter could be explained in part by degradation o f amino sugars but may also reflect the presence o f amidated derivatives o f dicarboxylic amino acids in native murein (Weiss et aL, 1968; Kato et aL, 1968). To see whether amidated amino acids were present, murein was treated with BTI (Soby and Johnson, 1981) which transforms carl3oxyamide groups into amines. The resulting derivatives showed altered retention times compared to the native compounds. Amino acid analysis of BTItreated T. neapolitanus murein revealed a strongly decreased DAP content. The molar ratio of Glu to DAP was 1:0.2, compared to 1:0.77 to 0.94 in untreated murein. A reduced DAP con-

J. BAJ E T AL.

50

retention time [min ]

Fig. 1. Amino acid and amino sugar analysis of a hydrolysate of T. neapolitanus murein. Murein was hydrolysedin 4 M HC1 under nit,ogen for 16 h at 105°C. The hydrolysatewas evaporated to dryness, washed several times with distilled water and re-evaporated. The final aqueous solution was passed through a 0.22-~.m "Millipore" filter prior to analysis with a "Durrum D-500" amino acid analyser.

tent was accompanied by the appearance o f an unidentified c o m p o u n d with a rentention time close to that of lysine. This result indicated that part o f the DAP in native murein is amidated.

Identification of compound X The retention time o f c o m p o u n d X indicated that i~ was more acidic than Mur (fig. 1). Compounds eluting before Mur were observed in hydrolysates of murein from B. cereus (Araki et al., 1972) and Chloroflexus aurantiacus (Jiirgens et al., 1987a) and were shown to be Mur-6P. To identify c o m p o u n d X from T. neapolitanus murein~ ~he products o f partial acid digestion were separated by cation exchange chromatography. Elution was first with water (fractions 1-11), then with 0.4 M HCI. Each fraction was

assayed for the presence o f hexosamines, neutral sugars and organic phosphorus. The results for the hexosamines are shown in figure 2. The fractions eluted with water contained a single amino sugar. The pooled material (fractions 1-3) was concentrated and acid hydrolyzed. It was found to contain 0.13 Izmol organic phosphorus per 0.i2 ~tmol hexosamine (molar ratio approx. 1/1). The amino sugar was colorimetrically identified as Mur. Thus c o m p o u n d X was Mur-P. A comparison o f the amount o f Mur-P and total Mur enabled us to calculate that approximately 1 out o f 24 residues o f MurNac is phosphorylated. Analysis o f the remaining fractions eluted with water, which did not contain hexosamines, revealed the presence o f organic phosphorus and an unidentified sugar. The molar ratio o f phos-

M U R E I N MODIFICATIONS I N THIOBACILLUS NEAPOLITANUS 1.5

',

',

,

,

i

.

.

.

.

i

r

~

,

51

To confirm the identity of c o m p o u n d Y in figure 1, murein was N-acetylated prior to acid hydrolysis and amino sugar and amino acid analysis. Under these conditions compound Y disappeared from the hydrolysate with a concomitant increase in the amounts c~f both GIcN and M u r (data not presented). In additional experiments, Y was found to be eluted from the amino acid analyser with an identical retention time as authentic GlcNI3(l-4)Mur, a kind gift from Dr. E. Ito.

,

.<

Dinitrophenylation of murein 0

10

20

30

fractions

Fig. 2. Chromatographic separation of amino sugars obtained as a result of mild hydrolysis of T. neapolitanus murein. Fractions 1 to 11 were eluted with water, fractions 12 to 25 with 0.4 M HCI. Assay for hexosamines was carried out to Tsuji et al. (1969). Each fraction was also tested for the presenceof orgaaic phosphorus and neutral sugars (dat~ not presented).

,',horus to sugar (calcula,.:.~ as salac:~ze cqui;'~ lents) was 1/2.25. it is therefore likely that the murein of T. neapolitanus contains an additional unidentified saccharide presumably linked t h r o u g h phosphate to C6 of Mur.

Murein was dinitrophenylated, hydrolysed in 4 M HC1 and analysed for the presence of substituted amino sugars by two-dimensional chromatography. Dinitrophenylated GlcN and M u r as well as hydrolysates of DNP-mureins from E. coli, T. versutus and B. cereus were used as st-~ndards. The results are presented in figure 3. The dilaitrophenylated sugar was identified as DNP-GIcN. The spot was scraped off the plate, eluted from the cellulose and its a m o u n t compared to DNP-GIcN standard. The acid hydrolysate of DNP-murein was also run in amino acid analyser. The molar ratio of GIcN to Glu was reduced, whereas that of ~,iur remained unaltered. Calculations based on the results of both methods revealed a lack of 30-35°70 N-acetyl substitution of GIcN in T. neapolitanus murein. No O-acetyl groups were found in the isolated murein.

Identification of compound Y The second unidentified c o m p o u n d (Y) e!uted very close to GIcN, in aome experiments it was even overlapped by the latter. A similar c o m p o u n d in an acid hydrolysate of B. cereus A H U 1356 murein was identified by Araki et al. (1972) as the disaccharide: GlcNl3(1-4)Mur. The occurrence o f this c o m p o u n d was found to be due to the presence of non-N-acetylated GIcN rcsidues in the murein. A similar compound was also obtained after acid hydrolysis of partly nonN-acetylated Rhodopseudomonas viridis murein (Schmelzer et aL, 1982).

Resistance of T. neapolitanus murein to lysozyme digestion Isolated murein from T. neapolitanus was poorly solubilized by lysozyme, as judged by changes in the absorbance of the murein suspension with time (fig. 4A). However, N-acetylation of murein made it a far better substrate. Absorbance of the suspension dropped close to 0 within 30 min of digestion. The behaviour of native and N-acetylated murein corresponded to that of native and similarly treated murein from

52

J. B A J E T AL.

A •

B o

0

A

|

0

T 1

2

3

t.

5

6 •

I.

E.coli

2.

T. ve~svh,l~

3.

T. neapolitanus

t,.

B. cereus AHU 1356

5.

DNP- glucosomine

6.

DNP- muramic acid

)

=

B

.'°°°

I

~.

~

5 -

~

Fig. 3. Thin layer chromatographyof acid hydrolysatesof dinitrophenylated mureins from T. neapolitanus, T. versums, E. coil and B. cereus. A = one-dimensionaiseparation, B = two-dimensional separation.

Chromatography was on cellulose F2545 plates in the followingsolventsystems: first direction, l-butanol pyridine water (6 4 3); second direction, 1.5 potassium phosphate buffer pH 6.0. DNP-compmmdswereseen as yellowspots.

B. cereus AHU1356. Better degradation o f both

mureins after N-acetylation was confirmed by a strong increase in reducing groups (fig. 4B) compared to native murcin. DISCUSSION

A pure murein fraction was isolated for the first time from T. neapolitanus and was found to contain only the characteristic constituents of murein of most other Gram-negative bacteria (Baj et al., 1991). However, the murein had a very !ow Ala content compared to other species. It was shown that this was not an artifact by analysis of E. coli and Caulobacter crescentus

4000

20

40

60

80

~00

120

t i m e (rain)

Fig. 4. Lysozymesensitivityof T. neapolitanus native (o) and N-acetylated(e) murein measured by a decreaseill absorbance (A) and an increase in reducing groups (B). A control experiment was run for native (~:) and Nacetylated (a) murein from B. cereus.

mureins performed under the same conditions (Markiewicz et al., 1983). Moreover, the low Ala content was confirmed by a strongly reduced amount of tetrapeptide-containing muropeptides in a muramidase digest o f T. neapolitanus murcin (Baj et al., 1991). This finding is not unc o m m o n and reflects the activity o f DD- and LD-carboxypeptidases (Markiewicz et aL, 1982). In the current study, murein o f T. neapolitanus was found to contain several further modifications which are infrequent in Gram-negative bacteria and very rarely occur in the murein o f a single species: (1) de-N-acetylation o f GlcN, (2) phosph~,rylation o f MurNac and (3) amidation o f DAP.

MUREIN MODIFICA 7t0NS IN THIOBACILLUS NEAPOLITANUS It seems that N-unsubstituted GlcN has so far been found in several species belonging to only two genera of Gram-negative bacteria (Schmelzer et aL, 1982; Jiirgens et al., 1987a). In the murein of T. neapolitanus 30-35~0 GlcN residues lack the N-acetyl group, which makes the maeromolecule a poor substrate for lysozyme. N-acetylation of murein greatly improved the lytie action of this enzyme. A second modification of the glycan strands found in murein of T. neapolitanus was the presence of Mur-P. Again, this modification is frequent in Gram-positive bacteria. Outside this group, MurNAc-6P has been found in the murein of certain cyanobacteria which seem to have developed a unique cell wall organization containing structural elements typical of both Gram-negative (outer membrane) and Grampositive (thick murein layer) bacteria (Jiirgens and Weckesser, 1986; Woitzik et aL, 1988). The position of the phosphate group on Mur in murein of T. neapolitanus has not been conclusively determined. Comparison of the results obtained for T. neapolitanus and B. cereus mureins shows that it is most probably on C6 and represents the r e m a i n d e r of a pyrophosphate linkage. This view would seem to be supported by our preliminary results indicating the presence of a saccharide presumably anchored (via a phosphodiester linkage) to Mur. Identification of the components of the saccharide and the nature of its linkage to murein are planned. The stem peptides of T. neapolitanus murein were found to contain amidated DAP residues. Amidation is frequent in murein from Grampositive bacteria and mainly concerns Glu residues. Amidated D A P has been found in murein of such species as B. subtilis, B. stearothermophilus and Corynebacterium diphtheriae (Rogers et aL, 1980). To our knowledge, T. neapolitanus is the first Gramnegative spe,:.:es whose murein carries this type of modification. Several Gram-negative species, whose murein shows structural variations, have been described by Weckesser's group. However, the bacteria in question, though belonging to different taxo-

53

nomic units, are all phototrophs. Prior to the analysis of mureins from Rhodopseudornonas viridis (Schmelzer et aL, 1982) or Rhodomicrobium vannieli (Jiirgens et al., 1987b) the murein of Gram-negative bacteria was not believed to show much structural variation. The modifications of T. neapolitanus murein described above may indicate that similar variations of murein exist in the large, but poorly characterized, group of chemolithotrophic bacteria, as is found among the phototrophs. The structure of murein has been suggested to be a useful chemotaxonomic marker (Schleifer and Seidl, 1985). Murein of 7". neapolitanus strongly differs from that of T. versutus, a facultatively chemolithotrophic species classified to the genus Thiobacillus (Baj and Markiewicz, 1988). Kelly and Harrisson (1989), based on the results of physiological and genetic studies, concluded that T. versutus was not related to obligatory autotrophic thiobacilli and that it should be withdrawn from the latter and classified in an appropriate genus of chemoorganotrophs. The difference in murein structure of T. versutus and T. neapolitanus, strongly supports the postulated lack of relatedness of both species.

Acknowledgements

The authors gratefully acknowledge the kind gifts of bacterial strains (Drs. J.-V. H61tje. O. q uovinen and E. lto) and murein from Bacillus cereus as well as disaccharide GIcNf(l-4)Mur (Dr. E. lto).

R~sidus glucosamine N-non-substitu~s et autres modifications dans la mur~ine de Thiobacillus neapolitanus chimiolithotrophe obligatoire

La mur6ine purihee de Thiobacillus neapolitam,s est peu dig~r6e par le lysozyme. Sa sensibilit~ b. l'er~zyme augmente beaucoup apr+s N-ac6tylation. Nous avons trouv6 que la mur~ine contient 30 a 35% de r~:sidus glucosamine sans groupe N-ac6tyl, et de l'acide phosphomuramique. Des modifications suppl6mentaires concernant (1) I'amidation de I'acide diamino-pim~lique dans les cha~nes lat6raies peptidiques et (2) un contenu faible en alanine. Aucune de ces modifications n'a ~t6 trouv6e dans ia mureine d'une autre bact6rie sulfo:euse, Thiobacillus versuIllS.

J. B A J E T A L . Mots-cl(s : Thlobacillus neapolitanus, Peptidoglycane, Glucosamine, Lysozyme; DAP, N-acdtyl, Chimiolithotrophie.

References Araki0 Y., Nakatal~i, T. & Nakayama, K. (1972), Occurrence of N-nonsubstituted glucosamine residues in peptidoglycan of lysozyme-resistant cell wall from Bacil&s cereus. J. biol. Chem., 247, 6312-6322. Baj, J. & Markiewicz, Z. (1988), Characterization of the cell wall murein of Thiobacillus versutus. Acta Microbiol. Polan., 37, 5-16. Baj, J. & Markiewicz, Z. (1991), The murein of Thiobacillus neapolitanus. Acta Microbiol. Polon., 40, 27-35. Glauner, B. (1988), Separation and quantification of muropeptides with high-performance liquid chromatography. Analyt. Biochem., 172, 451-464. Hadzija, O (1974), A simple method for the quantitative determination of muramic acid. Analyt. Biochem., 60, 512-517. Hayashi, H., Araki, Y. & Ito, E. (1973), Occurrence of glucosamine residues with free amino group~, in the cell wall peptidoglycan from bacilli as a factor responsible for resistance to lysozyme. J. Bact., 113,592-598. Hestrin, S. (1949), The reaction of aeetylcholine and other carboxylic acid derivatives with hydroxylamine and its analytical applic~:ion. J. biol. Chem., 180, 249 261. Jtirgens, U.J., Mcissner, J., Fischer, J., K6nig, W.A. attd Weckesser, J. (1987a), Ornithi-le as a constituent of the peptidoglycan of Chloroflexus aurantiacus, diaminopimelic acid in that of Chlorobium vibrioforme f. thiosulfatophilum. Arch. MKrobioL, 148, 72-76. Jiirgens, U.J., Rieth, B. & Wecke,.ser, J. (1987b), Partial lack of N-acetyl substitution of glueosamine in the pept[doglyean of the budding phototrophic Rhoaomicrobiurn vannielli. Z. NaturForsch., 42c, 1165-1170. Kato, K., Strominger, J.L. & Kotani, S. (1968), Structure of the cell wall of Corybebacterium diphtheriae. - I. Mechanism of h~clro!ysisby the L-3 enzyme and the structure of the peptide. Biochemistry, 2762-2772. Kelly, D.P. & Harrisson, A.P. (1989), Genus Thiobacillus, in "Bergey's Manual of Systematics Bacteriology" (J.G. Folt) (1842-1858) Williams & Wilkins, Baltimore. Lowry, O.H., Roberts, N.R., Leiner, K.Y., Wu, M.-J. & Farr, L. (1954), The quantitative histochemistry of the brain. Chemical methods. J. biol. Chem., 207, 1-17.

Markiewicz, Z., Glauner, B. & Schwarz, U. (1983), Murcin structure and the lack of DD- and LDcarboxypeptidase activities in Caulobacter crescentus. J. Baet., 156, 649-655. Markiewicz, Z., Broome-Smith, J.K., Schwarz, U. & Spratt, B.G. (1982), Spherical Escherichia coil due to elevated levels of D-alanine carboxypeptidase. Nature (Lond,), 297, 702-704. Miller, J.H. (1972), "Experiments in molecular genetics" (230-234). Cold Spring Harbor Laboratories, New York. Poindexter, J. & Hagenzieker, J.G. (1981), Constriction and septation during cell division in caulobacters. Canad. ,L MicrobioL, 27, 704-719. Rogers, H.J., Perkins, H.R. & Ward, J.B. (1980), Ultrastructure of bacterial envelopes, in "Microbial cell walls and membranes." Chapman & Hall, London, New York. Schleifer, K.H. & Kandler, O. (1972), Peptidoglycan types of bacterial cell walls and their taxonomic implications. Bact. Rev., 36, 407-477. Schleifer, K.H. & Stackebrandt, E. (1983), Molecular systematies of prokaryotes. Ann. Rev., 37, 143-187. Schleifer, K.H. & Seidl, P.H. (1985), Chemical composition and structure of murein, in "Chemical methods in bacterial systematics" (201-219). Society for Applied Bacteriology, New York. Schleifer, K.H. (1985), Analysis of the chemical com!3osition and primary structure of murein, in "Methods in microbiology" 18 (123-156). Academic Press, London, New York. Schmelzer, E., Weckesser, J., Warth, R. & Mayer, H. (1982), Peptidoglycan of Rhodopseudomonas viridis, partial lack of N-acetyl substitution of glucosamine. J. Bact., 149, 151-155. Soby, L.M. & Johnson, P. (1981), Determination of asparagine and glutamine in polypeptides using bis(l.l-trifluoroacetoxy)iodobenzene. Analyt. Biochem., 113, 149-153. Thompson, J.S. & Shockman, G.D. (1968), A modification of the Park and Johnson reducing sugar determination suitable for the assay of insoluble material: its application to bacterial cell walls. Analyt. Biochem., 22, 260-268. Tsuji, A., Kinoshita, T. & Hoshono, M. (1969), Microdetermination of hexosamines. Chem. Pharm. Bull., 17, 217-218. Vishniac, W. & Santer, M. (1957), The thiobacilli. Bact. Rev., 21, 195-213. Weiss, N., Plapp, R. & Kandler, O. (1967), Die Aminosauresequenz des DAP-haltigen Mureins yon Lactobacillus plantarum und Lactobacillus inulinus. Arch. MikrobioL, 58, 313-323. Woitzik, D., Weckesser, J. & Jiirgens, U.J. (1988), Isolation and characterization of cell wails components of the unicellular cyanobacterium Synechococcus sp. PCC6307. J. gen. Microbiol.., 134, 619-627.