Structure of the O-polysaccharide from Proteus myxofaciens Classification of the bacterium into a new Proteus O-serogroup Zygmunt Sidorczyk 1 , Anna N. Kondakova 2 , Krystyna Zych 1 , Sof’ya N. Senchenkova 2 , Alexander S. Shashkov 2 , Dominika Drzewiecka 1 and Yuriy A. Knirel 2 1 Department of General Microbiology, Institute of Microbiology and Immunology, University of Ło ´ dz ´ , Poland; 2 N.D. Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences, Moscow, Russian Federation The O-polysaccharide (O-antigen) was obtained from the lipopolysaccharide of Proteus myxofaciens,aProteus strain producing copious amounts of slime, which was isolated from the gypsy moth larvae. The structure of the polysaccharide was studied by chemical analysis and 1 H and 13 C NMR spectroscopy, including 2D COSY, TOCSY, ROESY and H-detected 1 H, 13 CHMQC experiments. It was found that the polysaccharide contains an amide of glucuronic acid (GlcA) with an unusual a-linked amino acid, N e -[(R)-1-carboxyethyl]- L -lysine (2S,8R-alaninolysine, 2S,8R-AlaLys), and has a linear tetrasaccharide repeating unit of the following structure: This structure is unique among known bacterial poly- saccharide structures. On the basis of these and sero- logical data, it is proposed that P. myxofaciens be classified into a new Proteus serogroup, O60, of which this strain is the single representative. Structural and serological relatedness of P. myxofaciens to other AlaLys- containing O-antigens of Proteus and Providencia is discussed. Keywords: lipopolysaccharide; N e -[(R)-1-carboxyethyl]- L -lysine; O-polysaccharide; O-serogroup; Proteus myxo- faciens. Gram-negative bacteria of the genus Proteus from the family Enterobacteriaceae are divided into four species: P. vulgaris, P. mirabilis, P. penneri and P. hauseri,aswell as three unnamed Proteus genomospecies 4, 5 and 6 [1,2]. They are widely distributed in nature and are important facultative human and animal pathogens, which in favorable conditions cause mainly intestine and urinary tract infections that sometimes lead to serious complica- tions, such as acute or chronic pyelonephritis and formation of bladder and kidney stones. They may also be the source of wound, burn, skin, nose, and throat infections [3]. Recently, it has been suggested that P. mirabilis plays an etiopathogenic role in rheumatoid arthritis [4]. Potential virulence factors of Proteus rods, which mediate infectious processes, are flagella, fimbriae, invasiveness, enzymes, e.g. proteases and ureases, haemolysins, capsular polysaccharide and lipopolysaccharide (endotoxin, LPS) [5,6]. The serological specificity of the bacteria is defined by the structure of the O-specific polysaccharide chain (O-antigen) of the lipopolysaccharide. On the basis of the O-antigens, the strains of two species, P. vulgaris and P. mirabilis, have been classified into 49 O-serogroups [7] and later into 11 additional serogroups [8]. About 15 further O-serogroups have been proposed for the third medically important species, P. penneri [9,10]. Structures of the O-polysaccharides of most Proteus serogroups have been determined and correlated with the immunospecificity of the O-antigens [9]. Most O-polysaccharides studied so far (>80%) contain acidic or both acidic and basic compo- nents, such as uronic acids, their amides with amino acids, phosphate, ethanolamine phosphate and other nonsugar constituents [9]. In 1966 another Proteus strain that produced copious amounts of slime was isolated from living and dead larvae of the gypsy moth (Porthetria dispar) and called Proteus myxofaciens [11]. Its medical importance and position in the serological classification of the genus Proteus remains unknown. In this work we report on the structure of the O-polysaccharide of P. myxofaciens and serological pro- perties of the LPS of this strain, including cross-reactivity with several other Proteus and Providencia strains with Correspondence to Z. Sidorczyk, Department of General Microbiology, Institute of Microbiology and Immunology, University of Ło ´ dz´ , Banacha 12/16, 90-237 Ło ´ dz´ , Poland. Fax: + 48 42 6784932, E-mail: zsidor@biol.uni.lodz.pl Abbreviations: EIA, enzyme immunosorbent assay; GlcA, glucuronic acid; HMQC, heteronuclear multiple-quantum correlation; LPS, lipopolysaccharide; 2S,8R-AlaLys and 2S,8S-AlaLys, N e -[(R)- and (S)-1-carboxyethyl]- L -lysine (2S,8R-alaninolysine and 2S,8S-alaninolysine). (Received 28 February 2003, revised 5 May 2003, accepted 2 June 2003) Eur. J. Biochem. 270, 3182–3188 (2003) Ó FEBS 2003 doi:10.1046/j.1432-1033.2003.03698.x structurally related O-polysaccharides. Together with the genera Proteus and Morganella, the bacteria Providencia comprise the third genus in the tribe Proteeae [12]. Members are distinguished by common morphological, cultural and enzymatic properties. Currently, the genus Providencia consists of five species: Pr. alcalifaciens, Pr. heimbache, Pr. rettgeri, Pr. rustigianii and Pr. stuartii [1,12]. On the basis of structural and serological data, we propose the classification of P. myxofaciens into a new, separate Proteus serogroup, O6O. Materials and methods Bacterial strains and growth P. myxofaciens strain 18769-CCUG (ATCC 19692) was kindly provided by E. Falsen [Culture Collection, Univer- sity of Goeteborg (CCUG), Goeteborg, Sweden]. P. mira- bilis O13 was from the Collection of the Institute of Microbiology and Immunology (University of Ło ´ dz´ ). Pr. alcalifaciens O23 and Pr. rustigianii O14 were from the Hungarian National Collection of Medical Bacteria (National Institute of Hygiene, Budapest, Hungary). Dry bacterial mass was obtained from aerated culture as described previously [13]. Isolation of the LPS and O-polysaccharide LPS of P. myxofaciens wasisolatedinayieldof6.23%of dried cell weight by phenol/water extraction [14] followed by treatment with cold aq. 50% trichloroacetic acid to precipitate nucleic acids as described [15]. The LPS was degraded with 1% (v/v) acetic acid at 100 °Cfor2h.The lipid precipitate was removed by centrifugation (13 000 g, 20 min), and the carbohydrate portion was fractionated by gel chromatography on a column (65 · 2.6 cm) of Sepha- dex G-50 using pyridinum acetate buffer as eluent (10 mL acetic acid and 4 mL pyridine in 1 L water) to give a high- molecular-mass O-polysaccharide in a yield of 20% of the LPS weight. Rabbit antisera and serological assays Polyclonal O-antisera were obtained by immunization of rabbits with heat-inactivated bacteria of P. myxofaciens and P. mirabilis O13 as described [16]. SDS/PAGE, electroblot- ting, immunostaining, enzyme immunosorbent assay (EIA), and absorption experiments were carried out as described [17]. LPS was used as antigen in the EIA. Sugar and amino-acid analyses Polysaccharide was hydrolysed with 2 M trifluoroacetic acid (120 °C, 2 h), and the monosaccharides were analysed by GLC as the acetylated alditols [18]. Amino components were identified using a Biotronik LC-2000 amino-acid analyzer on a column (0.4 · 22cm)ofOstionLGANB cation-exchange resin at 80 °Cin0.2 M sodium citrate buffer, pH 3.25, for amino acids and 0.35 M sodium citrate buffer, pH 5.28, for amino sugars. Uronic acid was identified using a Biotronik LC-2000 sugar analyzer on a Chromex DA · 8–11 column at 70 °Cin0.04 M KH 2 PO 4 buffer, pH 2.4. The absolute configuration of the amino sugars was determined by GLC of the acetylated (S)-2-butyl glycosides as described [19,20]. N e -[(R)-1-Carboxyethyl]- L -lysine was isolated from the polysaccharide hydrolysate (2 M trifluoroacetic acid, 120 °C, 2 h) by gel chromato- graphy on a column (80 · 1.6 cm) of TSK HW-40 in water. NMR spectroscopy 1 Hand 13 C NMR spectra were recorded with a Bruker DRX-500 spectrometer equipped with an SGI INDY computer workstation using internal acetone as reference (d H 2.225, d C 31.45). 2D NMR spectra were obtained using standard Bruker software, and the XWINNMR 2.1 program (Bruker) was used to acquire and process the NMR data. A mixing time of 200 and 300 ms was used in TOCSY and ROESY experiments, respectively. Results and discussion Structural studies LPS was obtained from dried bacterial cells of P. myxofac- iens and degraded with dilute acetic acid to give a high-molecular-mass O-polysaccharide isolated by gel-per- meation chromatography on Sephadex G-50. Analysis of the polysaccharide using an amino-acid analyzer revealed the presence of GlcN and GalN in the ratio 2 : 1 as well as another amino component. Sugar analysis using anion- exchange chromatography demonstrated the presence of glucuronic acid (GlcA). The D configuration of GlcN and GalN was established by GLC of the acetylated (S)-2-butyl glycosides. The D configuration of GlcA was determined by analysis of the 13 C NMR chemical-shift data of the polysaccharide (see below). The specific optical rotation values of 2S,8R-AlaLys for 2S,8S-AlaLys, +9.7 and +11.6, are too close to each other to be useful for differentiation between these two isomers, especially when the natural sample is not crystalline. However, the optical rotation value is useful for differen- tiation between 2S and 2R isomers because both 2S isomers (2S,8R and 2S,8S) have a positive value (published data [21]), whereas both 2R isomers (2R,8R and 2R,8S) would have a corresponding negative value, )9.7 and )11.6. Therefore, from the specific optical rotation value of +13 we can infer that the natural sample has the 2S configuration, i.e. that lysine is L as in all AlaLys isomers found so far in various natural sources. Differentiation between the 8S and 8R isomers in favor of the latter was made from the 13 C NMR data as described previously ([22,23]). 1 The 13 C NMR spectrum (Fig. 1) suggested that the polysaccharide is regular and has a tetrasaccharide repeat- ing unit. It contained signals for four anomeric carbons at d 98.7–104.0, five nitrogen-bearing carbons at d 47.0–58.8 (C2 of GlcN and GalN, C2 and C6 of AlaLys), one unsubsti- tuted (d 61.9) and two substituted (d 66.9 and 70.1) HOCH 2 -C groups (data of attached-proton test), one CH 3 -C group of AlaLys at d 16.2, three C-CH 2 -C groups of AlaLys at d 23.4, 26.5 and 31.9, three N-acetyl groups (CH 3 at d 23.5–23.8), and six carbonyl groups at d 170.1 (C6 of GlcA) and 175.2–177.9 (CO of N-acetyl groups and Ó FEBS 2003 O-Polysaccharide of P. myxofaciens (Eur. J. Biochem. 270) 3183 COOH of AlaLys). A relatively high-field position of the signal for C6 of GlcA showed that the uronic acid is amidated by AlaLys (compare with published data [23,24]). The 1 H NMR spectrum (Fig. 2) contained signals for four anomeric protons at d 4.40–4.92, one CH 3 -C group at d 1.47, three C-CH 2 -C groups at d 1.48–1.93, and three N-acetyl groups at d 2.01–2.10. These data together suggest that the tetrasaccharide repeating unit of the polysaccharide consists of two residues of GlcNAc and one residue each of GalNAc, GlcA and AlaLys. The 1 Hand 13 C NMR spectra of the polysaccharide were assigned using 2D NMR experiments, including 1 H, 1 H COSY, TOCSY, ROESY and H-detected 1 H, 13 C heteronuclear multiple-quantum correlation (HMQC), and the results are summarized in Table 1. On the basis of characteristic coupling constants [25], spin systems of two b-GlcpNAc residues (GlcNAc I and GlcNAc II ), a-GalpNAc and b-GlcpA, were identified. In particular, the configurations of the glycosidic linkages were deter- mined by the J 1,2 coupling constant values of 7–8 Hz for the b-linked and  4Hz for the a-linked sugar pyrano- sides. The remaining, nonsugar signals were assigned to AlaLys. The signal for H2 of AlaLys was shifted downfield to d 4.34, as compared with its position near d 3.8 in the free amino acid [23], indicating its acylation at N2. Low-field displacements of the signals for C4 of GlcA and C3 of b-GlcNAc II as well as C6 of b-GlcNAc I and a-GalNAc to d 77.6, 84.1, 66.9 and 70.1, as compared with their positions in the spectra of the corresponding unsub- stituted monosaccharides at d 72.9, 74.8, 61.9, and 62.1, respectively [26], demonstrated the modes of glycosylation of the monosaccharides. The linkage positions were confirmed and the sequence of the monosaccharides was determined using a 2D ROESY experiment, which revealed the following correlations between the anomeric protons and protons at the linkage carbons: GalNAc H1/GlcNAc I H6 at d 4.92/3.71 and 4.92/ 3.98, GlcNAc I H1/GlcpNAc II H3 at d 4.62/3.65, GlcNAc II H1/GlcA H4 at d 4.40/3.95, and GlcA H1/GalNAc H6 at d 4.58/3.87 and 4.58/4.01. In the 13 C NMR spectrum of the polysaccharide, a relatively large b effect ()2 p.p.m.) was observed on the C3 Fig. 1. 13 C NMR spectrum of the O-polysac- charide of P. myxofaciens. Fig. 2. 1 H NMR spectrum of the O-polysac- charide of P. myxofaciens. 3184 Z. Sidorczyk et al.(Eur. J. Biochem. 270) Ó FEBS 2003 signal of GlcA, which was caused by its glycosylation at position 4 with b- D -GlcNAc and showed that both mono- saccharides have the same absolute configuration, i.e. that GlcA has the D configuration (the b effect on C3 of L -GlcA wouldbeclosetozero[27]). On the basis of the data obtained, it was concluded that the repeating unit of the O-polysaccharide of P. myxofac- iens has the following structure: The polysaccharide contains an amide of D -glucuronic acid with an unusual amino acid, N e -[(R)-1-carboxyethyl]- L -lysine (structure 1 in Fig. 3). The same amide has previously been identified in the O-polysaccharide of Pr. alcalifaciens O23 [23], and an amide of the same amino acid with D -galacturonic acid (structure 2) in the O-polysac- charides of P. mirabilis O13 [24]. An amide of D -galacturonic acid with an isomeric amino acid, N e -[(S)-1-carboxyethyl]- L - lysine (structure 3), has been found in the O-polysaccharide of Pr. rustigianii O14 [28]. Structures of the O-polysac- charides that contain N e -[(R)-1-carboxyethyl]- L -lysine and N e -[(S)-1-carboxyethyl]- L -lysine are shown in Fig. 4. Some other O-polysaccharides of Proteus contain amide-linked L -lysine or aminoalkyl phosphate groups, which, like N e -(1-carboxyethyl)- L -lysine, endow the polysaccharides with a zwitterionic character. Serological studies LPS of a number of Proteus strains with known O-polysaccharide structure and those from Pr. rustigianii O14 and Pr. alcalifaciens O23 were tested for serological relatedness to P. myxofaciens. Only LPS of P. mirabilis O13 and Pr. rustigianii O14, both containing AlaLys [24,28], reacted strongly with polyclonal rabbit P. myxofaciens O-antiserum in an EIA, although the cross-reactivity was weaker than the reactivity of the homologous LPS (Table 2). P. mirabilis O13 O-antiserum reacted with the LPS of all three strains. However, the inhibiting doses of the cross-reactive LPS were significantly higher than that of the homologous LPS in both P. myxofaciens and P. mirabilis O13 homologous test systems (Table 2). Another AlaLys-containing LPS, that from Pr. alcalifaciens O23 [23], cross-reacted with O-antisera against P. myxofac- iens and P. mirabilis O13 only weakly (Table 2). Reactivity in EIA of both O-antisera with all tested LPS was completely abolished when they were absorbed with the homologous LPS (Table 3). Absorption of P. myxofaciens O-antiserum with the LPS of P. mirabilis O13 and Pr. rust- igianii O14 influenced its reactivity with the homologous LPS only slightly, whereas a significant decrease in the reactivity with the homologous LPS was observed when P. mirabilis O13 O-antiserum was absorbed with the LPS of P. myxofaciens and Pr. rustigianii O14. All tested LPS species completely removed cross-reactive antibodies from both O-antisera (Table 3). Absorption with Pr. alcalifaciens O23 LPS did not influence the reaction of the O-antisera with the other LPS species (data not shown). In Western blot, all tested LPS species showed similar patterns with O-antisera against P. myxofaciens and P. mir- abilis O13 with respect to slow moving LPS species with a long-chain O-polysaccharide (Fig. 5). Therefore, the cross-reactive LPS species share epitopes on the O-polysac- charide; the structures are shown in Fig. 4. The strong cross- reactivity of P. mirabilis O13 O-antiserum with the LPS of P. myxofaciens and Pr. rustigianii O14 (Table 2) is prob- ably due to the abundance of antibodies against AlaLys, which was demonstrated to be of importance in manifesting Table 1. 1 H and 13 C NMR chemical shifts for the O-polysaccharide of P. myxofaciens. Sugar residue Chemical shift (d, p.p.m.) Sugar atoms N-Acetyl or 1-carboxyethyl atoms H1 H2 H3 H4 H5 H6 H2¢ H3¢ fi 6)-a- D -GalpNAc-(1 fi 4.92 4.20 3.97 4.01 4.13 3.87 4.01 2.01 a fi 6)-b- D -GlcpNAc I -(1 fi 4.62 3.73 3.58 3.60 3.68 3.71 3.98 2.09 a fi 3)-b- D -GlcpNAc II -(1 fi 4.40 3.77 3.65 3.51 3.41 3.73 3.90 2.10 a fi 4)-b- D -GlcpA-(1 fi 4.58 3.36 3.63 3.95 3.97 2S,8R-AlaLys 4.34 1.76 1.93 1.48 1.72 3.06 3.72 1.47 C1 C2 C3 C4 C5 C6 C1¢ C2¢ C3¢ fi 6)-a- D -GalpNAc-(1 fi 98.7 50.9 68.5 69.7 71.0 70.1 175.2 a 23.5 b fi 6)-b- D -GlcpNAc I -(1 fi 102.7 56.8 74.8 70.7 75.3 66.9 175.6 a 23.5 b fi 3)-b- D -GlcpNAc II -(1 fi 100.6 55.2 84.1 70.0 76.7 61.9 175.7 a 23.8 b fi 4)-b- D -GlcpA-(1 fi 104.0 73.7 74.6 77.6 75.4 170.1 2S,8R-AlaLys 177.9 55.7 31.9 23.4 26.5 47.0 175.8 a 58.8 16.2 a,b Assignment could be interchanged. Ó FEBS 2003 O-Polysaccharide of P. myxofaciens (Eur. J. Biochem. 270) 3185 the P. mirabilis O13 specificity [29]. The configuration of neither AlaLys nor uronic acid amidated by AlaLys seems to play any significant role in the recognition. Indeed, antibodies against AlaLys could be completely absorbed by both LPS of Pr. rustigianii O14 containing AlaLys of a different configuration (2S,8S vs. 2S,8R) and that of P. myxofaciens containing a different uronic acid ( D -GlcA vs. D -GalA) (Table 3). These data are in full agreement with the results of serological studies with Pr. rustigianii O14 O-antiserum [28]. In contrast, P. myxofaciens O-antiserum predominantly contains antibodies to an epitope (or epitopes) different from AlaLys as it retained the reactivity with the homologous LPS after absorption with the LPS of P. mirabilis O13 or Pr. rustigianii O14 (Table 3). The lack of cross-reactivity of the Pr. alcalifaciens O23 LPS suggests that epitope(s) associated with AlaLys is feebly exposed on the LPS of this strain. Western blot (Fig. 5) also showed that both O-antisera tested contain antibodies that clearly recognized fast moving species of the homologous LPS and that these LPS bands of Pr. rustigianii O14 lack the O-polysaccharide chain and consist only of core and lipid A moieties. Therefore, epitopes on the core region, the structures of which remain unknown in all strains studied, may contribute to the cross- reactivity of the Pr. rustigianii O14 LPS. In summary, the structural and serological data show that P. myxofaciens possesses a unique O-antigen among Fig. 4. Structures of the O-polysaccharides of Proteus and Providencia containing N e -[(R)-1-carboxyethyl]- L -lysine and N e -[(S)-1-carboxy- ethyl]- L -lysine (2S,8R-AlaLys and 2S,8S-AlaLys, respectively). Table 2. EIA data of Proteus and Providencia LPS with rabbit poly- clonal O-antisera against P. myxofaciens and P. mirabilis O13. Data for the homologous LPS species are italicized. LPS Reactivity in EIA (reciprocal titer) Minimal inhibiting dose in the homologous test system in EIA (ng) P. myxofaciens O-antiserum P. myxofaciens O-antiserum/ P. myxofaciens LPS P. myxofaciens 1024 000 2 P. mirabilis O13 64 000 800 Pr. rustigianii O14 64 000 800 Pr. alcalifaciens O23 1600 P. mirabilis O13 O-antiserum P. mirabilis O-antiserum/ P. mirabilis LPS P. myxofaciens 256 000 125 P. mirabilis O13 512 000 1 Pr. rustigianii O14 256 000 62.5 Pr. alcalifaciens O23 2000 Fig. 3. Structures of amides of D -glucuronic and D -galacturonic acids with N e -[(R)-1-carboxyethyl]- L -lysine (1 and 2, respectively) and D -gal- acturonic acid with N e -[(S)-1-carboxyethyl]- L -lysine (3). 3186 Z. Sidorczyk et al.(Eur. J. Biochem. 270) Ó FEBS 2003 Proteus strains. We suggest it should be classified into a new Proteus serogroup, O60, of which this strain is the single representative. Acknowledgements This work was supported by the Science Research Committee (KBN, Poland; grant 6 P04 A 074 20), the Russian Foundation for Basic Research (grant 02-04-48767) and INTAS (grant YS 2001-2/1). References 1. O’Hara Mohr, C., Brenner, F.W. & Miller, J.M. (2000) Classifi- cation, identification and clinical significance of Proteus, Provi- dencia and Morganella. Clin. Microbiol. Rev. 13, 534–546. 2. O’Hara Mohr, C., Brenner, F.W., Steigerwalt, A.G., Hill, B.C., Holmes, B., Grimont, P.A.D., Hawkey, P.M., Penner, J.L., Miller, J.M. & Brenner, D.J. (2000) Classification of Proteus vulgaris biogroup 3 with the recognition of Proteus hauseri sp. nov. nom. re. & unnamed Proteus genomospecies. Int. J. Syst. Evol. Micro- biol. 50, 1869–1875. 3. Warren, J.W. (1996) Clinical presentation and epidemiology of urinary tract infections. In Urinary Tract Infections. Molecular Pathogenesis and Clinical Management (Mobley, H.T.L. & Warren,J.W.,eds),pp.2–28.ASMPress,Washington,DC. 4. Wilson, C., Thakore, A., Isenberg, D. & Ebringer. A. (1997) Correlation between anti-Proteus antibodies and isolation rates of P. mirabilis in rheumatoid arthritis. Rheumatol. Int. 16, 187–189. 5. Mobley, H.L.T. (1996) Virulence of Proteus mirabilis.InUrinary Tract Infections. Molecular Pathogenesis and Clinical Management (Mobley, H.T.L. & Warren, J.W., eds), pp. 245–269. ASM Press, Washington, DC. 6. Ro ´ z_ alski,A.,Sidorczyk,Z.&Kotełko, K. (1997) Potential viru- lence factors of Proteus bacilli. Microbiol. Mol. Biol. Rev. 61, 65–89. 7. Kauffmann, F. (1966) The Bacteriology of Enterobacteriaceae, pp. 333–360. Williams & Wilkins, Baltimore. 8. Penner, J.L. & Hennessy, J.N. (1980) Separate O-grouping schemes for serotyping clinical isolates of Proteus vulgaris and Proteus mirabilis. J. Clin. Microbiol. 12, 77–82. 9. Knirel, Y.A., Kaca, W.R., Ro ´ z_ alski, A. & Sidorczyk, Z. (1990) Structure of O-antigenic polysaccharides of Proteus bacilli. Pol. J. Chem. 73, 859–907. 10. Zych, K., Kowalczyk, M., Knirel, Y.A. & Sidorczyk, Z. (2000) New serogroups of the genus Proteus consisting of Proteus penneri strains only. Determination of some LPS epitopes responsible for their specificity. Adv. Exp. Med. Biol. 485, 339–344. 11. Cosenza, B.J. & Podgwaite, J.D. (1966) A new species of Proteus isolated from larvae of the gypsy moth Porthetria dispar. Antonie Van Leeuwenhoek Int. J. Microbiol. 32, 187–191. 12. Ewing, W.H. (1986) The tribe Proteeae: serologic reactions. In Identification of Enterobacteriaceae (Edwards, P.R., ed.), pp. 454–459. Elsevier, New York. 13. Koteł ko, K., Gromska, W., Papierz, M., Sidorczyk, Z., Kra- jewska-Pietrasik,D.&Szer,K.(1977)CoreregioninProteus mirabilis lipopolysaccharide. J. Hyg. Epidemiol. Microbiol. Immunol. 21, 271–284. 14. Westphal, O. & Jann, K. (1965) Bacterial lipopolysaccharides. Extraction with phenol-water and further applications of the procedure. Methods Carbohydr. Chem. 5, 83–91. 15. Zych, K., Toukach, F.V., Arbatsky, N.P., Kołodziejska, K., Senchenkova, S.N., Shashkov, A.S., Knirel, Y.A. & Sidorczyk, Z. (2001) Structure of the O-specific polysaccharide of Proteus mir- abilis D52 and typing this strain to Proteus serogroup O33. Eur. J. Biochem. 268, 4346–4351. 16. Zych, K., S ´ wierzko, A. & Sidorczyk, Z. (1992) Serological char- acterization of Proteus penneri species novum. Arch. Immunol. Ther. Exp. 40, 89–92. Fig. 5. Western blot of Proteus and Providencia LPS with P. myxo- faciens (A) and P. mirabilis O13 (B) O-antisera. P. myx. refers to P. myxofaciens,PmO13toP. mirabilis O13, and Pr O14 to Pr. rust- igianii O14. Table 3. EIA data with rabbit polyclonal O-antisera against P. myxo- faciens and P. mirabilis O13 absorbed with Proteus and Providencia LPS. Data for the homologous LPS species are italicized. Sheep red blood cells were used as control. Alkali-treated LPS used for absorption Reactivity (reciprocal titer) with LPS from: P. myxofaciens P. mirabilis O13 Pr. rustigianii O14 P. myxofaciens O-antiserum Control 1024 000 64 000 64 000 P. myxofaciens <1000 <1000 <1000 P. mirabilis O13 512 000 <1000 <1000 Pr. rustigianii O14 512 000 <1000 <1000 P. mirabilis O13 O-antiserum Control 256 000 512 000 256 000 P. myxofaciens <1000 32 000 <1000 P. mirabilis O13 <1000 <1000 <1000 Pr. rustigianii O14 <1000 32 000 <1000 Ó FEBS 2003 O-Polysaccharide of P. myxofaciens (Eur. J. Biochem. 270) 3187 17. Sidorczyk, Z., Zych, K., Toukach, F.V., Arbatsky, N.P., Zabłotni, A., Shashkov, A.S. & Knirel, Y.A. (2002) Structure of the O-polysaccharide and classification of Proteus mirabilis strain G1 in Proteus serogroup O3. Eur. J. Biochem. 269, 1406–1412. 18. Sawardeker, J.S., Sloneker, J.H. & Jeanes, A. (1965) Quantitative determination of monosaccharides as their alditol acetates by gas liquid chromatography. Anal. Chem. 37, 1602–1603. 19. Gerwig, G.J., Kamerling, J.P. & Vliegenthart, J.F.G. (1979) Determination of the absolute configuration of monosaccharides in complex carbohydrate by capillary G.L.C. Carbohydr. Res. 77,1–7. 20. Shashkov, A.S., Senchenkova, S.N., Nazarenko, E.L., Zubkov, V.A., Gorshkova, N.M., Knirel, Y.A. & Gorshkova, R.P. (1997) Structure of a phosphorylated polysaccharide from Shewanella putrefaciens strain S29. Carbohydr. Res. 303, 333–338. 21. Fujioka, M. & Tanaka, M. (1978) Enzymic and chemical synthesis of e-N-( L -propionyl)- L -lysine. Eur. J. Biochem. 90, 297–300. 22. Thompson, J. & Miller, S.P.F. (1988) N 6 -(1-Carboxyethyl) lysine formation by Streptococcus lactis.Purification,synthesis,and stereochemical structure. J. Biol. Chem. 263, 2064–2069. 23. Kocharova, N.A., Vinogradov, E.V., Borisova, S.A., Shashkov & Knirel, Y.A. (1998) Identification of N e -[(R)-1-carboxyethyl]- L -lysine in, and the complete structure of, the repeating unit of the O-specific polysaccharide of Providencia alcalifaciens O23. Carbohydr. Res. 309, 131–133. 24. Perepelov, A.V., Senchenkova, S.N., Cedzynski, M., Zio ´ łkowski, A.,Vinogradov,E.V.,Kaca,W.,Shashkov,A.S.&Knirel,Y.A. (2000) Isolation using triflic acid solvolysis and identification of N e -[(R)-1-carboxyethyl]-N a -( D -galacturonoyl)- L -lysine as a com- ponent of the O-specific polysaccharide of Proteus mirabilis O13. Carbohydr. Res. 328, 441–444. 25. Altona, C. & Haasnoot, C.A.G. (1980) Prediction of anti and gauche vicinal proton–proton coupling constants in carbohy- drates: a simple additivity rule for pyranose rings. Org. Magn. Reson. 13, 417–429. 26. Jansson, P E., Kenne, L. & Widmalm, G. (1989) Computer- assisted structural analysis of polysaccharides with an extended version of CASPER using 1 H- and 13 C-n.m.r. data. Carbohydr. Res. 188, 169–191. 27. Shashkov, A.S., Lipkind, G.M., Knirel, Y.A. & Kochetkov, N.K. (1988) Stereochemical factors determining the effects of glycosy- lation on the 13 C chemical shifts in carbohydrates. Magn. Reson. Chem. 26, 735–747. 28. Kocharova, N.A., Zatonsky, G.V., Torzewska, A., Macieja, Z., Bystrova, O.V., Shashkov, A.S., Knirel, Y.A. & Ro ´ z_ alski, A. (2003) Structure of the O-specific polysaccharide of Providencia rustigianii O14 containing N e -[(S)-1-carboxyethyl]-N a -( D -galac- turonoyl)- L -lysine. Carbohydr. Res. 338, 1009–1016. 29. Swierzko, A.S., Cedzynski, M., Zio ´ łkowski, A., Senchenkova, S.N., Perepelov, A.V., Knirel, Y.A. & Kaca, W. (2001) Structure and serological characterization of an N e -[(R)-1-carboxyethyl)- L -lysine-containing O-chain of the lipopolysaccharide of Proteus mirabilis O13. Arch. Immunol. Ther. Exp. 49, 163–169. 3188 Z. Sidorczyk et al.(Eur. J. Biochem. 270) Ó FEBS 2003 . of the polysaccharide consists of two residues of GlcNAc and one residue each of GalNAc, GlcA and AlaLys. The 1 Hand 13 C NMR spectra of the polysaccharide were. copious amounts of slime was isolated from living and dead larvae of the gypsy moth (Porthetria dispar) and called Proteus myxofaciens [11]. Its medical importance and