Structure of the O-polysaccharide leads to classification of<i>Proteus penneri</i>31 in<i>Proteus</i>serogroup O19
Bibliographic record
Abstract
Being common in human and animal intestines, under favourable conditions Gram-negative bacteria of the genus Proteus from the family Enterobacteriaceae cause various infections. From them, most important are urinary tract infections, which can lead to severe complications, such as acute or chronic pyelonephritis and formation of bladder and kidney stones. After Proteus mirabilis and Proteus vulgaris, Proteus penneri is the third distinguished species of these medically important bacteria. As opposite to the first two species, serological classification of P. penneri is far from being complete. Recently, on the basis of immunochemical studies of the lipopolysaccharide (LPS), a number of new O-serogroups have been proposed for P. penneri strains [1]. O-polysaccharides of most P. penneri strains studied are acidic due to the presence of uronic acids and various acidic non-sugar components, such as amino acids, lactic and pyruvic acids and phosphate groups [2–5]. Now we report on the structure of another acidic O-polysaccharide isolated from P. penneri strain 31, which shares some structural features with the O-polysaccharides of several other Proteus strains studied by us earlier. Serological relatedness of the LPS of some of these strains was revealed and served as the basis for classification of P. penneri 31 in Proteus serogroup O19. P. penneri strains 16 (CDC 465-75), 20 (CDC 0766–80), 28 (ITJ 339) and 31 (TGH 629, isolated from urine of a patient with bacteriuria in Toronto, Canada) were kindly provided by Prof. D.J. Brenner (Center for Diseases Control, Atlanta, GA, USA). Strains of P. vulgaris and P. mirabilis were from the Czech National Collection of Type Cultures (CNCTC, Institute of Microbiology and Genetics, Prague, Czech Republic). Dry bacterial cells of P. penneri 31 were obtained from aerated liquid culture as described [6]. LPS was obtained by extraction of dried bacterial cells with a hot phenol–water mixture [7] and purified by treatment with cold aqueous 50% CCl3CO2H followed by dialysis of the supernatant [8]. Degradation of LPS was performed with aqueous 1% HOAc at 100°C for 2 h, and products were fractionated by gel-permeation chromatography on a column (3×65 cm) of Sephadex G-50 in 0.05 M pyridinium acetate buffer (pH 4.5) monitored using a Knauer differential refractometer (Germany) to give a high-molecular-mass polysaccharide. Polyclonal O-antiserum was obtained by immunisation of rabbits with heat-inactivated bacteria of P. penneri 31 according to the published procedure [9]. SDS–PAGE, electrotransfer of LPS from gels to nitrocellulose sheets, immunostaining and absorption experiments were carried out as described [10]. LPS was used as antigen in enzyme immunosorbent assay (EIA) [11]. Passive immunohemolysis (PI) was performed with increasing amounts (2–200 µg) of alkali-treated LPS [11]. The polysaccharide was hydrolysed with 2 M CF3CO2H (120°C, 3 h). Amino components were identified using a Biotronik LC-2000 amino acid analyser equipped with a column (22×0.4 cm) of Ostion LG AN B cation-exchange resin in 0.2 M sodium citrate buffer (pH 3.25) at 65°C. Neutral sugars were analysed using a Biotronik LC-2000 sugar analyser as described [12]. The alditol acetates were derived as described [13] and analysed by gas-liquid chromatography (GLC) using a Hewlett-Packard 5890 chromatograph (USA) equipped with a DB-5 fused-silica capillary column. The absolute configurations of the monosaccharides were determined by GLC of acetylated (+)-2-octyl glycosides according to the published method [14] modified as described [15]. Methylation was performed according to the Hakomori procedure [16]. Prior to methylation, the polysaccharide was dephosphorylated with aqueous 48% HF (4°C, 16 h). The methylated polysaccharide was recovered by extraction with ethyl acetate, hydrolysed as in sugar analysis, partially methylated sugars were conventionally converted into alditol acetates and analysed by GLC-MS using a Hewlett-Packard 5890 chromatograph equipped with a NERMAG R10–10L mass spectrometer (France) under the same chromatographic conditions as in sugar analysis. 1H, 13C, and 31P NMR spectra were recorded with a Bruker DRX-500 spectrometer in D2O at 60°C using internal acetone (δH 2.225, δc 31.45) or external aqueous 85% H3PO4 (δP 0) as reference. 2D NMR spectra were obtained using standard Bruker software, and the XWINNMR 2.1 program (Bruker) was used to acquire and process the data. A mixing time of 150 and 300 ms was used in 2D TOCSY and ROESY experiments, respectively. Other parameters used for 2D experiments were essentially the same as described previously [17]. The O-polysaccharide was obtained by mild acid degradation of the LPS of P. penneri 31 followed by gel chromatography on Sephadex G-50. Chemical analyses using amino acid and sugar analysers after hydrolysis of the polysaccharide revealed galactose and four amino components: 2-amino-2-deoxyglucose, 2-amino-2-deoxygalactose, 2-amino-2,6-dideoxygalactose (FucN) and 2-aminoethanol in almost equal amounts. The identification of the monosaccharides was additionally confirmed by GLC of the alditol acetates and acetylated (+)-2-octyl glycosides, which showed the l configuration of FucN and the d configuration of the other monosaccharides. The 13C-NMR spectrum of the polysaccharide demonstrated a tetrasaccharide repeating unit. The spectrum showed signals for four anomeric carbons at δ 98.7–103.9, two unsubstituted HOC H2-C groups (C6) at δ 61.6 and 61.9, one O-substituted HOC H2-C group at δ 65.5 (data of attached-proton test [18]), one C H3-C group at δ 16.2 (C6 of FucN), three nitrogen-bearing carbons at δ 49.6, 51.2 and 56.8 (C2 of amino sugars), 13 sugar ring oxygen-bearing carbons in the region δ 67.8–80.7, three N-acetyl group (CH3 at δ 23.4–23.5, CO at δ 174.8–175.3) and one residue of 2-aminoethanol (CH2N at δ 41.5 and CH2O at δ 62.1). Accordingly, the 1H-NMR spectrum of the polysaccharide contained signals for four anomeric protons at δ 4.67–5.12, one CH3-C group at δ 1.19 (doublet, J5,6 6.26 Hz, H6 of FucN), three N-acetyl groups at δ 1.96–2.03 (all singlets) and one residue of 2-aminoethanol (CH2N at δ 3.28, triplet, J 5 Hz). The 31P-NMR spectrum of the polysaccharide contained one signal for a monophosphate group at δ 3.2. Therefore, the polysaccharide has a tetrasaccharide repeating unit that contains one residue each of d-Gal, d-GlcNAc, d-GalNAc, l-FucNAc and 2-aminoethyl phosphate (EtnP). Methylation analysis using GLC-MS of the partially methylated alditol acetates derived from the dephosphorylated polysaccharide demonstrated 3-substituted Gal, GlcNAc and FucNAc, and 4-substituted GalNAc. Therefore, the polysaccharide is linear. The 1H- and 13C-NMR spectra were assigned using 2D shift-correlated experiments, including 1H,1H COSY, TOCSY, ROESY and H-detected 1H,13C HMQC, and the chemical shifts are tabulated in Tables 1 and 2. The sugar residues were identified as Glcp NAc (J3,4 and J4,5∼10 Hz), Galp NAc, Fucp NAc and Galp (J3,4 and J4,5≤3 Hz). Fucp NAc was recognised by a H5/H6 correlation at δ 4.42/1.19 in the COSY spectrum and a H5/H4 correlation at δ 4.42/3.91 in the ROESY spectrum, as well as by a low-field position of the signal for H2 at δ 4.32 and its correlation to the signal for C2 at δ 49.6 in the 1H,13C HMQC spectrum. Similarly, the spin systems of Glcp NAc and Galp NAc were distinguished by correlation of the signals for the protons at nitrogen-bearing carbons (H2) to the corresponding carbons (C2) at δ 56.8 and 51.2. H NMR data of the O-polysaccharide of P. penneri 31 (δ, ppm) Chemical shifts for NAc are δ 1.96, 1.98 and 2.03. C NMR data of the O-polysaccharide of P. penneri 31 (δ, ppm) Chemical shifts for NAc are δ 23.4, 23.4 and 23.5 (all Me), 174.8, 175.0 and 175.3 (all CO). As judged by a relatively large J1,2 coupling constant value of 7–8 Hz for the H1 signal at δ 4.67, GlcNAc is β-linked. The signals for the three remaining anomeric protons were not clearly resolved; their relatively low-field position at δ 4.93–5.12 enabled suggestion that they are α-linked. The configurations of the glycosidic linkages were confirmed by the ROESY spectrum, which showed correlations between H1 and H3,5 for the β-linked GlcNAc and between H1 and H2 for the remaining, α-linked sugar residues. The 1H,31P HMQC experiment showed correlations of the phosphorus signal at δ 3.2 with the proton signals of the 2-aminoethanol group at δ 3.28 (CH2N) and 4.12 (CH2O), as well as with the H6 signal of GlcNAc at δ 4.17. These data indicated that EtnP is attached at position 6 of GlcNAc, which was confirmed by a low-field position of the signal for C6 of GlcNAc at δ 65.5 (α-effect of phosphorylation; compare the chemical shift δ 61.9 for C6 in the non-substituted monosaccharide [21]). Low-field displacements in the 13C-NMR spectrum of the signals for C3 of Gal, GlcNAc and FucNAc and C4 of GalNAc by 4–11 ppm (α-effect of glycosylation), as compared with their positions in the spectra of the corresponding unsubstituted monosaccharides [19], revealed the glycosylation pattern of the monosaccharide residues in the repeating unit, which was in agreement with the methylation analysis data (see above). In addition to intraresidue NOE correlations, the ROESY spectrum of the polysaccharide showed the following interresidue correlations between the anomeric protons and protons at the linkage carbons: Gal H1/GalNAc H4 at δ 4.93/4.09; GalNAc H1/FucNAc H3 at δ 5.12/4.02; FucNAc H1/GlcNAc H3 at δ 5.01/3.73; and GlcNAc H1/Gal H3 at δ 4.67/3.95. These data are in agreement with the glycosylation pattern and defined the full sequence of the monosaccharide residues in the repeating unit. Therefore, the O-polysaccharide of P. penneri 31 has the structure 1 shown in Fig. 1. Structures of the O-polysaccharides of P. penneri 31 (1) (this work), P. vulgaris O19 (2) [23], P. penneri 16 and 20 (3) [24] and P. penneri 28 (4) [25]. d-Fuc3N(R-3HOBu) stands for 3,6-dideoxy-3-[(R)-3-hydroxybutyramido]-d-galactose and d-GlcNAc3(S-Lac) for 2-acetamido-3-O-[(S)-1-carboxyethyl]-2-deoxy-d-glucose (N-acetylisomuramic acid). This polysaccharide has some structural features in common with several other Proteus O-polysaccharides. For instance, β-d-GlcNAc6P Etn has been found in the O-polysaccharides of P. mirabilis O27 [20], P. penneri 8 (O67) [21] and P. penneri 63 (O68) [22]. Furthermore, an α-l-FucNAc-(1→3)-d-GlcNAc fragment is present in all Proteus polysaccharides that contain FucNAc [2,21–23]. Most remarkably, the O-polysaccharide of P. penneri 31 has the same carbohydrate backbone as the O-polysaccharide of P. vulgaris O19 [23] (structure 2 in Fig. 1), from which it differs in the presence of EtnP only. From the LPS of 130 Proteus strains tested, only five reacted with rabbit polyclonal P. penneri 31 O-antiserum in PI. These were the homologous LPS and those of P. penneri 16, 20, 28 and P. vulgaris O19 (Table 3; for the O-antigen structures of these strains [23–25] see Fig. 1). The last strain reacted as strong as the homologous LPS, whereas the reactivity of P. penneri 28 was slightly weaker and that P. penneri 16 and 20 much weaker. Similar results were obtained in EIA (Table 3). Almost no cross-reaction was observed with the LPS of P. mirabilis O27, P. penneri 8 (O67) and P. penneri 63 (O68), which share with P. penneri 31 a β-d-GlcNAc6P Etn residue (all three strains) and an α-l-FucNAc-(1→3)-d-GlcNAc disaccharide (P. penneri 8 and P. penneri 63) [20–22]. Reactivity of O-antisera against P. penneri 31 and P. vulgaris O19 with the Proteus LPS LPS and alkali-treated LPS were used as antigen in EIA and PI, respectively. The data of the homologous LPS are shown in bold type. The specificity of the cross-reactions was confirmed by inhibition experiments in the system P. penneri 31 O-antiserum/P. penneri 31 LPS using PI and EIA (Table 3). The LPS of P. penneri 31 and P. vulgaris O19 strongly inhibited the reaction, whereas the other heterologous LPS were significantly weaker inhibitors. In Western blot (Fig. 2A), P. penneri 31 O-antiserum reacted with slow migrating bands of the homologous LPS and the LPS of P. vulgaris O19, P. penneri 16 and 20 as well as with fast migrating bands of all tested antigens. The slow and fast migrating bands correspond to the LPS species consisting of a core-lipid A moiety with or without a O-polysaccharide chain attached to the core, respectively. The slow migrating band pattern of the O-polysaccharide-containing LPS species of P. penneri 31 LPS was similar to that of P. vulgaris O19 and different from those of P. penneri 16 and 20. The fast migrating band patterns were similar for P. penneri 16, 28 and 31 and different for P. penneri 20 and P. vulgaris O19. Absorption of P. penneri 31 O-antiserum with the LPS of P. penneri 28 removed antibodies that recognise both O-polysaccharide and core epitopes in the LPS of P. penneri 16, 20 and 28 but did not affect binding pattern of the LPS of P. penneri 31 and P. vulgaris O19 (data not shown). Western blot of Proteus LPS with O-antisera against P. penneri 31 (A) and P. vulgaris O19 (B). P.p. and P.v. stand for P. penneri and P. vulgaris, respectively. The reactivity of P. penneri 31 O-antiserum with all tested antigens in PI was completely abolished when it was absorbed with the homologous LPS (Table 4). Absorption with the LPS of P. vulgaris O19 removed antibodies to this LPS, decreased the reactivity with the P. penneri 31 LPS and did not influence the reaction with P. penneri 16, 20 and 28. Absorption with each of the last three LPS had no influence on the reaction with the P. vulgaris O19 LPS, slightly decreased the reactivity with the P. penneri 31 LPS and completely abolished that with the three other antigens. Double absorption first with the P. penneri 28 LPS to remove anti-core antibodies and then with the P. vulgaris O19 LPS to remove antibodies against a common O-polysaccharide epitope(s) decreased the reactivity with the homologous LPS and completely abolished the reactivity with all heterologous antigens. The remaining antibodies are thus directed against an epitope(s) on the O-polysaccharide of P. penneri 31 that is absent from P. vulgaris O19. PI of the alkali-treated Proteus LPS with absorbed O-antisera against P. penneri 31 and P. vulgaris O19 Sheep red blood cells were used as control. Studies with P. vulgaris O19 O-antiserum using PI and EIA as well as inhibition experiments (Table 3) confirmed close serological relatedness between P. vulgaris O19 and P. penneri 31. Western blot (Fig. 2B) showed a similar binding pattern of P. vulgaris O19 O-antiserum to slow migrating LPS bands of both strains and a different binding pattern to fast migrating bands. These data suggest similarity in O-polysaccharide epitopes and both similarity and differences in core epitopes of P. vulgaris O19 and P. penneri 31. Absorption of P. vulgaris O19 O-antiserum with the homologous LPS completely abolished the reactivity with both antigens, whereas absorption with the P. penneri 31 LPS kept minor antibodies reactive with the homologous LPS (Table 4), which, most likely, are core-specific. P. vulgaris O19 O-antiserum did not cross-react with the LPS of P. penneri 16, 20 and 28 in any of the assays. As mentioned above, the O-polysaccharide of P. penneri 31 has the same carbohydrate backbone as that of P. vulgaris O19 [23], the two antigens differing only in the absence of EtnP from the latter (Fig. 1). The structural and serological data of the LPS suggest that it is reasonable to classify P. penneri 31 to Proteus O19 serogroup and divide this serogroup to two subgroups: O19a for P. vulgaris O19 and O19a,19b for P. penneri 31. Factor O19a is linked to a common epitope on the O-polysaccharides and factor O19b to a particular epitope that is present on the O-polysaccharide of P. penneri 31 only and is evidently associated with EtnP. O19b seems to be a minor epitope since sharing by the O-polysaccharides of not only EtnP but also GlcNAc6P Etn could not provide any significant cross-reactivity of the corresponding LPS with P. penneri 31 O-antiserum. This finding suggests that the corresponding antibody may recognise a larger epitope, including GlcNAc6P Etn and a neighbouring sugar. Previously, the importance of EtnP and its N-[(R)-1-carboxyethyl] derivative (d-AlaEtnP) in manifesting the serological specificity has been demonstrated for some other Proteus O-antigens. For instance, the presence of EtnP in the O-antigen of P. mirabilis O17 is one of the factors that are responsible for its serological distinctions from the O-antigen of P. penneri 20 and some other cross-reactive P. penneri strains, which was the basis for classification of the P. mirabilis and P. penneri strains in one Proteus serogroup O17 as two different subgroups [24]. On the other hand, two serologically related P. mirabilis strains, PrK 28/57 and EU313, have been classified into the same Proteus serogroup O14 based on serological data and the presence in their O-antigens of α-d-Galp 6(d-AlaEtnP) as the only common component [26]. This work was supported by the Russian Foundation for Basic Research (grant 02-04-48767), INTAS (grant YS 2001–2/1) and the Sciences Research Committee (KBN, Poland, grant 6 P04A 074 20). enzyme immunosorbent assay 2-aminoethyl phosphate lipopolysaccharide 2-acetamido-2,6-dideoxygalactose (N-acetylfucosamine) passive immunohemolysis
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.002 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".