INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

The genus Acinetobacter (Moraxellaceae [26]) is a group of gram-negative coccobacilli which can be found widespread in nature (18, 32), though most strains have also been isolated from various samples of animal and human origin (8, 11, 18, 27, 28). In general, Acinetobacter strains are not virulent (2, 18); however, some of them are involved in severe infections of immunocompromised patients in intensive care units with increasing frequency (2, 11, 31, 32). The infections are often difficult to treat, since many of the isolates are highly resistant to antibiotics (1, 2, 31). However, the importance of these organisms in hospitals may be underestimated, due to difficulties in the identification of Acinetobacter strains to the species level (2, 5, 7, 35). As a result of DNA-DNA hybridization studies, 20 DNA homology groups (genomic species) have been delineated, 7 of which have received formal species names (3, 4, 9, 30). Though Acinetobacter baumannii (genomic species 2) is the most prevalent species associated with outbreaks of nosocomial infection (2, 11, 31, 32), strains belonging to the unnamed genomic species 13 Sensu Tjernberg and Ursing (13TU) (30) have also been isolated from a number of outbreaks (6, 29). Unfortunately, phenotypic investigations of large numbers of Acinetobacter strains, validated by DNA homology studies, have shown that some DNA groups can only be identified unambiguously by DNA-DNA hybridization techniques (10). This has been shown to be particularly problematic for species belonging to the A. calcoaceticus-A. baumannii complex (5, 10), which also includes the clinically relevant genomic species 2 and 13TU. However, DNA-DNA hybridization is laborious and time-consuming, making it inadequate for use in clinical microbiology laboratories (5). Other molecular identification methods have been proposed, but most are either too expensive, require too much experience or standardization, or have simply proven to be unsatisfactorily discriminative (5, 12). Thus, there is still a need for simple, reliable, and inexpensive identification methods for Acinetobacter strains, especially for those belonging to the above-mentioned clinically relevant DNA groups, methods that can be implemented in clinical microbiology laboratories (5).

Acinetobacter, like other gram-negative bacteria, contains lipopolysaccharide (LPS) on the surface of its outer membrane, which in many strains has been shown to be of the smooth (S) phenotype (13-17, 33, 34). Therefore, we are investigating the possibility of an identification scheme for Acinetobacter strains based on the different O antigens within the genus, as has been done for other bacterial genera (19, 21, 24, 25). For this purpose, we have started the generation of monoclonal antibodies (MAbs) against the O-antigenic polysaccharide from different Acinetobacter isolates of clinical and environmental origin.

We report here on the generation and characterization of an MAb specific for the O-antigen moiety of the LPS of strains belonging to the unnamed genomic species 13TU, and we evaluate its use as part of an O-serotyping scheme for the identification of these bacteria at the species level.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacteria. The Acinetobacter strains of genomic species 13TU investigated in this study (Table 1) consisted of a selection of representative isolates associated with outbreaks in Germany and in Denmark, as well as a number of nonoutbreak strains. Additional Acinetobacter strains, belonging to other genomic species, were also examined (Table 2). All isolates had been identified previously to the species level by a wide variety of methods, including DNA-DNA hybridization, pulsed-field gel electrophoresis (PFGE) analysis, ribotyping, biotyping, and electrophoretic protein profiling (6, 29). Epidemiologically related isolates had identical biotype and ribotype patterns, but in some cases different PFGE typing patterns were observed (29). The non-Acinetobacter strains used in this study were obtained from R. Podschun (National Reference Center of Klebsiella species, Kiel, Germany) or were from our own culture collection (Table 2). All bacteria were preserved in 10% (vol/vol) glycerol broth at 70°C until further use.

Bacterial LPS, whole-cell lysates, and proteinase K digestion. LPS was extracted from Acinetobacter sp. strain 108 as described previously (33). Preparation of whole-cell lysates (undiluted or diluted 1:4 in sample buffer [33]) and proteinase K digestion were performed as reported previously (23). The samples were used immediately or otherwise stored at 20°C. In the latter case, they were heated again at 100°C for 5 min prior to use.

MAb. The MAb was prepared by conventional protocols after immunization of mice with heat-killed bacteria. Acinetobacter sp. strain 108, against which rabbit hyperimmune sera had been produced in a previous report (23), was selected in this study for immunization. Four BALB/c mice were injected intravenously on days 0, 7, 14, and 21 with 20, 20, 60, and 120 µg of antigen, respectively. The mice were boosted intravenously on day 125 and intraperitoneally on days 126 and 127 with 200 µg of antigen each, followed by fusion on day 129. Primary hybridomas were screened by dot blot and enzyme immunoassay (EIA) with LPS as antigen. The relevant hybridoma was cloned three times by limiting dilution, isotyped with a commercially available isotyping kit (Bio-Rad), and purified by affinity chromatography on Protein G (Pharmacia). Purity was subsequently checked by Coomassie staining after sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and stored at 20°C.

Serological methods. EIA and dot and Western blot analyses were performed as described earlier (23, 33), with purified LPS or proteinase K-treated bacterial lysates as antigens.

Acid hydrolysis of membrane-bound LPS. Membrane-bound LPS was hydrolyzed in 0.1 M HCl as described in a previous study (22), with minor modifications. Briefly, undiluted bacterial lysates were treated with proteinase K, subjected to SDS-PAGE, and subsequently transferred overnight onto a polyvinylidene difluoride (PVDF) membrane, which was then incubated at 100°C for 1 h in a heat-resistant glass container containing 0.1 M HCl. After an extensive washing in blot buffer (33), the membrane was blocked in blot buffer supplemented with 10% nonfat dry milk and immunostained with lipid A-specific MAb S1 (20) as described elsewhere (22). A parallel SDS-polyacrylamide gel was stained with alkaline silver nitrate as described previously (33).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Immunization of mice and preparation of MAb. BALB/c mice were successfully immunized with heat-killed bacteria from Acinetobacter sp. strain 108. Animals were tested on day 28 for serum antibodies against the antigen used for immunization by dot blot assay. The animal exhibiting the strongest reactivity was used for fusion on day 129. Primary hybridomas (n = 864) were tested by dot blot and by EIA for antibody reactivity, with LPS as antigen; 10 of these produced specific antibodies. One was finally selected for further studies on the basis of good reactivity in EIA and high specificity in the dot blot assay. The antibody (S48-13) was cloned three times by limiting dilution and was found to be of the immunoglobulin G1 class. The MAb was purified by using protein G; purification was ascertained by SDS-PAGE and Coomassie staining (data not shown). The results described below were obtained with the affinity-purified antibody.

Specificity of MAb in EIA. MAb S48-13 was tested by EIA, with LPS as antigen. Reactivity with the homologous antigen was observed at a concentration of 2.5 ng of antibody per ml (data not shown). In addition, checkerboard titrations were performed, with antigen concentrations of between 32 and 4,000 ng/ml (1.6 to 200 ng of antigen per well) and antibody concentrations of between 0.5 and 1,000 ng/ml. The results (Fig. 1) showed that the antibody exhibits good reactivity with the homologous LPS over a broad range of antigen concentrations. The specificity of the antibody could be visualized by Western blot (see below).

View larger version (24K): [in this window] [in a new window]   FIG. 1.   Checkerboard titration of MAb S48-13 in EIA with native LPS from Acinetobacter sp. strain 108 as a solid-phase antigen. Plates were coated with antigen concentrations of 4,000 (), 2,000 (), 1,000 (), 500 (), 250 (), 125 (), 63 (), and 32 () ng/ml of coating solution (50 µl/well) and reacted with MAb at the concentrations indicated on the abscissa. OD405, optical density at 405 nm.

Reactivity of MAb in dot and Western blots. When proteinase K-digested whole-cell lysate or LPS from Acinetobacter sp. strain 108 was separated by SDS-PAGE, blotted onto a PVDF membrane, and immunostained with MAb S48-13, a banding pattern typical of an O-polysaccharide chain could be observed (Fig. 2, lane 1). No staining was observed with the core lipid A portion when proteinase K-treated lysate was separated on a 15% gel (data not shown), thus indicating that MAb S48-13 does not react with the core oligosaccharide portion of the LPS molecule and that the observed banding pattern is not the result of a core-reactive antibody. To confirm that MAb S48-13 is indeed directed against the O antigen and not another polysaccharide proteinase K-treated bacterial lysate, which after SDS-PAGE had been immobilized on a PVDF membrane, was hydrolyzed in 0.1 M HCl, and the membrane-bound 4'-monophosphoryl lipid A was subsequently detected in situ with lipid A-specific MAb S1. A ladder-pattern, indistinguishable from that observed following immunostaining with MAb S48-13, could be observed following visualization of bands with MAb S1 (data not shown), thus indicating that the antibody is indeed directed against the O polysaccharide.

View larger version (94K): [in this window] [in a new window]   FIG. 2.   Reactivity of MAb S48-13 in a Western blot with clinical isolates of unnamed genomic species 13TU after separation of the proteinase K-treated whole-cell lysates (10 µl each) by SDS-PAGE on a 10% separating gel. The bacterial lysates are in the following lanes: 1, strain 108; 2, strain 353; 3, strain 387; 4, strain 4419; 5, strain 9894; 6, strain 9836; 7, strain 10716; 8, strain 10717; 9, strain 12112; 10, strain 53937bb; 11, strain 3417; 12, strain 3418; 13, strain 3419; 14, strain 3420; 15, strain 3421; 16, strain St-11681; 17, strain St-7961; 18, strain St-8195; and 19, strain St-2312.

View larger version (56K): [in this window] [in a new window]   FIG. 3.   Reactivity of MAb S48-13 with clinical isolates of unnamed genomic species 13TU in dot blot. Proteinase K-treated bacterial lysates were diluted 1:3 in distilled water, dotted onto a nitrocellulose membrane (1 µl per dot), and immunostained. The bacteria are as follows (from left to right): top row, strains 108, 353, 387, 4419, 9894, 9836, 10716, 10717, and 12112; second row, strains 53937bb, 3417, 3418, 3419, 3420, 3421, St-11681, St-7961, and St-8195; and third row, strain St-2312.

Determination of LPS phenotype by silver staining. Proteinase K-treated bacterial lysates from the strains which had not reacted with MAb S48-13 were subjected to SDS-PAGE, and the gel was subsequently stained with alkaline silver nitrate. As shown in Fig. 4A, a distinct ladder pattern could be observed for strains 353 (lane 1), 383 (lane 2), 53937bb (lane 4), 3420 (lane 8), and 3421 (lane 9). However, no banding pattern was observed for the other strains.

View larger version (86K): [in this window] [in a new window]   FIG. 4.   Determination of LPS phenotype after SDS-PAGE on a 10% gel and staining with alkaline silver nitrate (A) or with MAb S1 in a Western blot after hydrolysis at 100°C in 0.1 M HCl (B) of proteinase K-treated bacterial lysates (15 µl each) from strains which had not reacted with MAb S48-13. Lanes: 1, strain 353; 2, strain 387; 3, strain 4419; 4, strain 53937bb; 5, strain 3417; 6, strain 3418; 7, strain 3419; 8, strain 3420; 9, strain 3421.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Due to lack of phenotypic methods for the rapid and unambiguous identification of Acinetobacter strains at the species level (5, 10), we are currently investigating the possibility of an identification for the genus Acinetobacter based on the O-antigenic polysaccharide of the LPS from these bacteria. Using hyperimmune rabbit sera, we could show that such a scheme is feasible (23). However, though they were shown to be highly specific, the antisera have the disadvantage of containing core-reactive and non-LPS antibodies, e.g., capsular and protein antibodies, which would lead to false-positive results when the sera are used for identification purposes (23). MAbs, however, can be generated which are O antigen specific and thus are more suitable for such a scheme. Moreover, MAbs have the advantage of being producible in virtually unlimited amounts.

In this study, a murine MAb was generated against Acinetobacter strains belonging to the unnamed genomic species 13TU (30) which, next to genomic species 2 (A. baumannii), is the most frequent DNA group associated with nosocomial infections in intensive care units (2, 31). The antibody, S48-13, was characterized by EIA and Western blot analysis and was found to be highly specific for the O-antigenic polysaccharide of the homologous LPS. It has been shown in previous studies that Acinetobacter sp. strain 108 produces S-type LPS (22, 33), and this could be confirmed in our study. MAb S48-13 was additionally tested by dot and Western blot with 18 other clinical isolates belonging to unnamed genomic species 13TU, which had been characterized previously by several methods, including DNA-DNA hybridization, biotyping, ribotyping, and PFGE. Although epidemiologically related strains were observed to have identical bio- and ribotype patterns (6, 29), in some cases differing PFGE types were noted (29). Reactivity was observed with four outbreak and five nonoutbreak strains. A different O-antigen banding pattern was observed for three of the nonoutbreak strains from Denmark which had reacted positively with MAb S48-13. This is due to the presence of a chemically different but antigenically related O antigen in the LPS of these strains compared to that of the other isolates, since the migration distance of the core lipid A region of these strains after SDS-PAGE was the same as that of the homologous strain (data not shown). This will be confirmed by structural analysis.

To determine their LPS phenotype, those strains which had not reacted with MAb S48-13 were subjected to an acid hydrolysis procedure (22) in which, after SDS-PAGE and transfer of the LPS onto a PVDF membrane, the membrane is incubated in an acidic solution at 100°C for 1 h, resulting in the liberation of lipid A which remains membrane bound and can be detected in situ by using a lipid A-specific MAb such as S1 (20). Surprisingly, the ladder patterns observed for these strains were identical, suggesting that these strains may represent an additional serotype within genomic species 13TU. The sensitivity of this method is much greater than that of the silver-staining procedure and has been discussed in detail elsewhere (22). When SDS-polyacrylamide gels of proteinase K-digested lysates of the strains which did not react with MAb S48-13 were stained with alkaline silver nitrate, banding patterns were observed for five strains. However, with the acid hydrolysis method, O-antigen characteristic patterns could be observed for these five strains as well as for three additional strains. One isolate, strain 3419, did not show a distinct banding pattern, which is suggestive of a reduced O-antigen expression or the natural production of LPS of the rough phenotype.

MAb S48-13 did not react with any Acinetobacter strains belonging to other genomic species, including the closely related DNA groups 1, 2, and 3, or with the LPS of strains from several other gram-negative bacteria, such as Escherichia coli, Salmonella spp., Shigella sonnei, Enterobacter spp., Klebsiella pneumoniae, Pseudomonas spp., Stenotrophomonas maltophilia, Serratia spp., and Proteus spp. Moreover, only one additional putative serotype was found within the group of isolates that were investigated, suggesting that Acinetobacter genomic species 13TU may possibly only consist of a few serotypes. Thus, a mixture of a few MAbs may become a convenient tool to definitely identify strains belonging to Acinetobacter DNA group 13TU.

Although the antibody described in this study does not identify all clinical isolates, the data show that genomic species 13TU may be, due to its relatively low degree of diversity, particularly suited for a serotyping scheme with MAbs. To provide a more comprehensive understanding of the antigenic variations in Acinetobacter O antigens, we will perform the structural analysis of those antigens which are not detected by the antibody described here and generate new MAbs to fill the present gap. However, we also need the cooperation of clinical microbiologists willing to evaluate these antibodies in various clinical settings in different geographic areas. One should keep in mind that the well-established serotyping schemes for Salmonella, Shigella, E. coli, and other gram-negative human pathogens are extremely helpful tools and were not established at once.