Lipopolysaccharide (endotoxin, LPS) is an important potential virulence factor of Proteus rods. The serological specificity of the bacteria is defined by the structure of the O-polysaccharide chain (O-antigen) of the LPS. Until now, 76 O-serogroups have been differentiated among Proteus strains.
Materials and Methods:
LPSs were isolated from Proteus mirabilis TG 83, TG 319, and CCUG 10700 (OA) strains by phenol/water extraction. Antisera were raised by immunization of rabbits with heat-killed bacteria. Serological investigations were performed using enzyme immunosorbent assay, passive immunohemolysis, inhibition of both assays, absorption of antisera, and Western blot.
Results:
The cross-reactive epitope shared by these strains and P. penneri O72a,O72b is located on the O-polysaccharide and is most likely associated with an α-D-Glcp-(1→6)-β-D-GalpNAc disaccharide fragment. The serological data indicated the occurrence of two core types in the LPSs studied, one characteristic for P. mirabilis TG 319 and CCUG 10700 (OA) and the other for P. mirabilis TG 83 and O57.
Conclusions:
The serological and structural data showed that P. mirabilis TG 83, TG 319, CCUG 10700 (OA), and O57 have the same O-antigen structure and could be qualified to the Proteus O57 serogroup.
Abkürzungen
EIA
enzyme immunosorbent assay
LPS
lipopolysaccharide
PIH
passive immunohemolysis
CCUG
Culture Collection of the University of Goeteborg
Introduction
Gram-negative bacteria of the genus Proteus are widely distributed in the environment, such as soil, water, and sewage, and represent a part of the normal microflora of human and animal intestines. They are opportunistic pathogens which in favorable conditions cause intestine and urinary tract infections that may lead to serious complications, such as acute or chronic pyelonephritis, the formation of bladder and kidney stones, and catheter obstruction [15, 19]. In addition, Proteus strains may be a source of wound, burn, skin, nose, and throat infections [10]. P. mirabilis also plays an etiopathogenic role in rheumatoid arthritis [21]. The genus Proteus is currently divided into five species, including P. vulgaris, P. mirabilis, P. penneri, P. hauseri, and P. myxofaciens, as well as three unnamed Proteus genomospecies 4, 5, and 6 [10, 11].
Lipopolysaccharide (endotoxin, LPS) is an important potential virulence factor of Proteus [14]. The serological specificity of the bacteria is defined by the structure of the O-polysaccharide chain (O-antigen) of the LPS. Based on the O-antigens, strains of two species, P. vulgaris and P. mirabilis, were originally classified into 49 O-serogroups [2]. Our serological studies combined with structural elucidation of the O-polysaccharides resulted in reclassification of some of the existing 49 O-serogroups and the creation of 18 new O-serogroups (O52, O56, O58-O73) for the third medically important species, P. penneri, as well as P. hauseri, P. myxofaciens, and Proteus genomospecies [1, 15, 24, 25]. Similar studies were also performed with P. mirabilis and P. vulgaris strains from the collections of Penner and Hennessy [12] and Larsson et al. [9]. As a result, about half of them (O8, O23, O34, O40, O65, O69) were classified into the existing serogroups and a further six serogroups (O50, O53, O54, O55, O74, O75) were proposed for the others [3‐6, 13, 22, 23].
Anzeige
Here we report the serological characterization of the LPSs of three unclassified P. mirabilis strains, TG 83, TG 319 [12], and CCUG 10700 (OA) [9], all of which are candidates for the Proteus serogroup O57.
Materials And Methods
Bacterial strains and growth
Proteus mirabilis TG 83 and TG 319 were kindly provided by Prof. J. L. Penner (Department of Medical Genetics, University of Toronto, Canada). P. mirabilis OA strain CCUG 10700 as well as P. myxofaciens strain CCUG 18769 were obtained from the Culture Collection of the University of Goeteborg, Goeteborg, Sweden. Twenty-four strains of P. penneri came from the Collection of the Department of General Microbiology, University of Łódź, Poland. The P. hauseri strain was kindly provided by C. M. O'Hara and D. J. Brenner (Centers for Disease Control and Prevention, Atlanta, Georgia, USA). P. vulgaris (27) and P. mirabilis (39) strains were from the Czech National Collection of Type Cultures (CNCTC, National Institute of Public Health, Prague, Czech Republic). The bacteria were cultivated under aerobic conditions on nutrient broth (BTL, Łódź, Poland). Dry bacterial mass was obtained as described elsewhere [8].
Isolation and degradation of the LPS
LPSs were obtained from bacterial cells by hot phenol/water extraction [20] and purified using a cold aqueous 50% CCl3CO2H precipitation procedure as described elsewhere [27].
Rabbit antisera and serological assays
Polyclonal P. mirabilis OA, TG 319, and TG 83 O-antisera were obtained by immunization of rabbits with heat-inactivated bacteria according to a published procedure [26]. Passive immunohemolysis test (PIH), enzyme immunosorbent assay (EIA), inhibition of these two tests, as well as absorption experiments were carried out as described elsewhere [17]. SDS/PAGE, electrotransfer of the LPS from gels to nitrocellulose sheets, and immunostaining were performed according to a published procedure [17].
Anzeige
Results and Discussion
In order to reveal a possible serological relatedness of P. mirabilis TG 83, TG 319, and CCUG 10700 (OA) to each other and to other Proteus strains, O-antisera against these three strains were tested in PIH and EIA with the LPSs of 94 Proteus strains, including 40 P. mirabilis, 27 P. vulgaris, and 24 P. penneri strains as well as one strain each of P. myxofaciens, P. hauseri and Proteus genomospecies 4. Only five LPSs were crossreactive, namely those of P. mirabilis TG 83, TG 319, CCUG 10700, O57 and, in addition, P. penneri 4 (O72a,72b). The data of the tests and the amount of each antigen necessary for the inhibition of the reactions in PIH and EIA are given in Table 1.
Table 1
Reactivity of O-antisera against P. mirabilis TG 83, TG 319, and CCUG 10700 (OA) strains with the Proteus LPSa, b
LPS from
Reciprocal titer for the LPS
Minimal inhibitory dose (ng) of the LPS
PIH
EIA
PIH
EIA
P. mirabilis TG 83 O-antiserum
P. mirabilis TG 83
51 200
512 000
2
2
{tiP. mirabilis} TG 319
51 200
512 000
2
2
P. mirabilis CCUG 10700
51 200
512 000
2
2
P. mirabilis O57
51 200
512 000
2
2
P. penneri O72a,72b
6 400
64 000
250
500
P. mirabilis TG 319 O-antiserum
P. mirabilis TG 83
25 600
256 000
2
2
P. mirabilis TG 319
25 600
256 000
2
2
P. mirabilis CCUG 10700
25 600
256 000
2
2
P. mirabilis O57
25 600
256 000
2
2
P. penneri O72a,72b
3 200
32 000
500
1 000
P. mirabilis CCUG 10700 (OA) O-antiserum
P. mirabilis TG 83
51 200
512 000
2
2
P. mirabilis TG 319
51 200
512 000
2
2
P. mirabilis CCUG 10700
51 200
512 000
2
4
{tiP. mirabilis} O57
51 200
512 000
2
2
P. penneri O72a,72b
3 200
8 000
500
500
aLPS and alkali-treated LPS were used as antigen in EIA and PIH, respectively.
bData for homologous LPS are italicized.
The strongest cross-reactivity with all O-antisera was observed for the LPSs of P. mirabilis TG 83, TG 319, and CCUG 10700 and O57. It was on the same level as the reactivity of the homologous LPS, suggesting serological identity of all four strains. The LPS of P. penneri O72a,72b showed a weaker cross-reactivity and a weaker inhibiting activity when tested in the homologous systems with all the O-antisera used.
In Western blot (Fig. 1), all O-antisera clearly recognized slow migrating bands of both homologous and heterologous LPS, and the banding patterns of P. mirabilis TG 83, TG 319, CCUG 10700, and O57 were almost identical. These bands correspond to high-molecular-mass LPS species with O-polysaccharides consisting of a large number of repeating units, and hence the cross-reactive epitope(s) is located on the O-polysaccharide. The recognition by heterologous O-antisera of fast migrating bands of P. mirabilis TG 83 and O57 LPS, corresponding to the core-lipid A moiety, was different from that of TG 319 and CCUG 10700. No binding was observed for the low-molecular-mass LPS species of P. penneri O72a,72b.
For a more detailed epitope characterization, the antigens were tested in PIH with O-antisera after absorption with various LPSs (Table 2). The reactivity of the O-antisera with all the tested antigens was completely abolished when they were absorbed with the homologous LPS. Absorption of O-antisera with LPS from either P. mirabilis TG 319 or CCUG 10700 removed all cross-reactive antibodies against the other LPS of the set. In contrast, absorption with P. mirabilis TG 83 and O57 LPS of O-antisera against P. mirabilis TG 319 or CCUG 10700 left a small fraction of antibodies specific to the homologous LPS (titer 1:800/1600). This finding and the Western blot data (see above) allowed the suggestion that this fraction recognized an epitope(s) in the LPS core region, which seems to be different in TG 83, O57, and two other strains.
Table 2
Passive immunohemolysis data of alkali-treated LPS with absorbed O-antisera against P. mirabilis TG 83, TG 319, and CCUG 10700 (OA) strainsa, b
O-antisera absorbed with the alkali-treated LPS from
Reciprocal titer of absorbed with the alkali-treated LPS from
P. mirabilis
P. penneri
TG 83
TG 319
CCUG 10700
O57
O72a,72b
P. mirabilis TG 83 O-antiserum
Control
51 200
51 200
51 200
51 200
6 400
P. mirabilis TG 83
<100
<100
<100
<100
<100
P. mirabilis TG 319
800
<100
<100
800
<100
P. mirabilis CCUG 10700
800
<100
<100
800
<100
P. mirabilis O57
<100
<100
<100
<100
<100
P. penneri O72a,72b
12 800
12 800
12 800
12 800
<100
P. mirabilis TG 319 O-antiserum
Control
25 600
25 600
25 600
25 600
3 200
P. mirabilis TG 83
<100
800
800
<100
<100
P. mirabilis TG 319
<{ur100}
<100
<100
<100
<100
P. mirabilis CCUG 10700
<100
<100
<100
<100
<100
P. mirabilis O57
<100
<100
<100
<100
<100
P. penneri O72a,72b
12 800
12 800
12 800
12 800
<100
P. mirabilis CCUG 10700 O-antiserum
Control
51 200
51 200
51 200
51 200
3 200
P. mirabilis TG 83
<100
1 600
1 600
<100
100
P. mirabilis TG 319
<100
<100
<100
100
100
P. mirabilis CCUG 10700
<100
<100
<100
<100
<100
P. mirabilis O57
<100
<100
<100
<100
<100
P. penneri O72a,72b
12 800
12 800
12 800
12 800
<100
aSheep red blood cells were used as a control.
bData for homologous LPS are italicized.
Absorption of all O-antisera with the LPS of P. penneri O72a,72b decreased the level of antibodies from the titer 1:25,600/51,200 to 1:12,800, thus confirming the presence in all the LPSs studied of an epitope common to the P. penneri O72a,72b LPS. As revealed by the Western blot data (see above), this epitope is located in the O-polysaccharide region.
The serological results were in agreement with the structural analysis of the O-polysaccharides isolated from the LPS of P. mirabilis TG 83 and TG 319. Comparison of their 13C-NMR spectra with each other and with that of P. mirabilis ATCC 49995 (serogroup O57) studied earlier [18] using the “finger-print” method showed that they are essentially identical, and hence the three O-polysaccharides have the same structure shown in Fig. 2 (for the assignment of the 1H- and 13C-NMR spectra, see published data [18]). Based on the combined serological and structural data, we consider P. mirabilis TG 83, TG 319, and CCUG 10700 (OA) as candidates for the Proteus O57 serogroup as the next three representatives.
A comparison of the O-polysaccharide structures (Fig. 2) allowed the suggestion that the cross-reactive epitope of P. penneri O72a,72b [7, 16] is associated with an α-D-Glcp-(1→6)-β-D-GalpNAc disaccharide fragment shared by all the strains studied.
In conclusion, 1) the serological and structural data showed that P. mirabilis TG 83, TG 319, CCUG 10700 (OA), and O57 have the same O-antigen structure and may be considered candidates for the Proteus O57 serogroup; 2) the cross-reactive epitope shared by these strains and P. penneri O72a,72b is located on the O-polysaccharide and is most likely associated with an α-D-Glcp-(1→6)-β-D-GalpNAc disaccharide fragment; 3) the serological data pointed to the occurrence of two core types in the LPSs studied, one characteristic for P. mirabilis TG 319 and CCUG 10700 (OA) and the other for P. mirabilis TG 83 and O57.
Anzeige
×
×
Acknowledgment
This work was supported by the State Committee for Scientific Research (KBN, Poland, grant No. 2 PO5A 086 26) and the Council on Grants of the President of the Russian Federation for the Support of Young Russian Scientists (project No. MK-2204.2006.4). The authors thank Janusz Włodarczyk for excellent technical assistance in the preparation of the manuscript.
Open Access This is an open access article distributed under the terms of the Creative Commons Attribution Noncommercial License (
https://creativecommons.org/licenses/by-nc/2.0
), which permits any noncommercial use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.