TEXT

Top Abstract Text References

Burkholderia cepacia, originally described as a plant pathogen (3), has become an important opportunistic pathogen in patients with cystic fibrosis (CF) (10, 11, 14, 26) and an infrequent cause of nosocomial infection in patients without CF (13). There is increased evidence of transmission among patients with CF by social contact (9) The environment is also regarded as a potential source of strains capable of infecting these patients (4, 9). However, one risk factor for B. cepacia acquisition by patients with CF in the United States appeared to be hospitalization, and a recent hospital outbreak apparently involved patients both with and without CF (13). There are no clear data about the possible contribution of the polysaccharide produced extracellularly by specific B. cepacia isolates (1, 5, 6, 20) to the colonization and persistence of the species in the infected host, as was ascribed to alginate in the respiratory infection of patients with CF by Pseudomonas aeruginosa (10). Nevertheless, the mucoid colonial morphotype is thought to be relatively rare among B. cepacia CF isolates (10). However, the exopolysaccharides (EPSs) produced by several gram-negative bacterial species infecting plants or animals have been considered important virulence factors due to their contributions to the colonization and persistence of the producing microorganism in the infected host (19).

In Portugal, the first clearly identified B. cepacia isolate recovered from the sputum of a patient with CF attending the CF treatment center at Sta. Maria Hospital in Lisbon was found in 1992. The CF Center is attended by approximately 85% of the population with CF residing in the Lisbon area and by patients with CF living in the south of Portugal and in the Madeira and Azores Islands. From 9 of a total of 140 patients with CF registered at the CF Center between 1995 and the end of 1997, 23 isolates capable of growing on the selective Burkholderia Cepacia Selectatab medium (Mast Diagnostics, Merseyside, United Kingdom) were recovered at the Laboratory of Bacteriology of Sta. Maria Hospital. The isolates (Table 1) were obtained on different dates of isolation from bronchial secretions of different patients with CF, each identified by a letter (Table 1), after 3 days of incubation at 35°C followed by another day of incubation at room temperature in the selective medium Burkholderia Cepacia Selectatab. The four putative B. cepacia isolates obtained during 1996 were lost due to a prolonged storage period at the hospital and could not be examined in this study. To confirm that all 19 CF isolates obtained belonged to the species B. cepacia, the commercial systems API 20NE (Biomerieux, Marcy L'Etoile, France) and BIOLOG gram-negative (GN) (Biolog Inc., Hayward, Calif.) were used. The isolates identified as B. cepacia by the two systems were submitted to additional confirmation based on PCR amplification using the specific oligonucleotide primers CMG-16-1, C-16-2₁₀₀₁, CMG-23-1, CM-16-2₄₅₈, and G-23-2 proposed by Bauernfeind et al. (2). Further efforts to confirm the identification of the 19 isolates were undertaken at the Instituto Superior Técnico (IST) laboratory before comparing their genomic relatedness and their abilities to produce EPS, due to the increasing evidence of misidentification of B. cepacia by standard laboratory procedures (2). This analysis revealed that only the 16 IST isolates listed in Table 1 could be confirmed to be B. cepacia. They were stored at 70°C in 40% (wt/vol) glycerol, and when in use, the cultures were routinely maintained on Pseudomonas isolation agar (Difco, Detroit, Mich.) plates.

B. cepacia prevalence rate and clinical course. The value of the 3-year cumulative prevalence rate found during the period of surveillance (6.4%) did not suggest epidemic transmission of B. cepacia. This value is close to the prevalence rate observed in other surveillance studies carried out in other countries (11, 21, 22, 25), while the colonization rates in one large United Kingdom regional center, experiencing the spread of an epidemic strain, reached values close to the 40% prevalence experienced in a major North American CF center (9). Consistent with previous reports, the clinical courses of the nine Portuguese patients with CF infected with B. cepacia who were followed during this study were variable: patients A, D, and I exhibited stable pulmonary function after infection with B. cepacia; patient G died from the "cepacia syndrome"; patients B and H showed an increased deterioration of pulmonary function which was clearly associated with infection with the same B. cepacia isolate that continued to be isolated during 1998, and patients E and F died, although their deaths were considered unrelated to B. cepacia infection. Patient C moved to another geographical area, which prevented his clinical observation during the full duration of the present study.

Molecular typing. The genetic relatedness of the 16 B. cepacia isolates from Portuguese patients with CF was compared by ribotyping complemented with macrorestriction fragment analysis by pulsed-field gel electrophoresis (PFGE). The environmental type strain ATCC 25416 and the three highly transmissible epidemic strains J2315 (a representative of the highly epidemic Edinburgh-Toronto lineage [11]) and C1394 and C1579 (epidemic representatives of outbreaks of B. cepacia among patients with CF attending CF centers at Manchester [23] and Glasgow [27], United Kingdom, respectively) were used as reference strains. For ribotyping analysis, the isolation of total DNA, restriction with EcoRI (Gibco BRL Life Technologies, Gaithersburg, Md.), DNA blotting, and hybridization with the acetylaminofluorene-labeled 16S + 23S rRNA probe from E. coli (Eurogentec, Seraing, Belgium) were carried out as previously described (16). The sizes of the hybridization restriction fragments were estimated with DNA Simdex software from GenetX with RaoulI (Eurogentec) as the reference molecular size marker. In order to transform the restriction fragment length polymorphism (RFLP) data obtained from ribotyping for numerical analysis, each band of an RFLP profile obtained for each isolate under study was treated as a unit character and scored as 1 when present or 0 when absent across all isolates. Variations in staining intensity were not taken into account for the construction of the matrix. From the resulting binary data, a triangular similarity matrix was generated using the Dice similarity coefficient, S_D (24). The construction of the dendrogram from the similarity matrix was performed by the UPGMA method (unweighted-pair group method using arithmetic means), which forms groups by successively pairing similar ribopatterns according to the magnitudes of their observed distances (24). The software package used was the program NTSYSpc version 2.02 (Exeter Software, Inc.). The cophenetic correlation, r, between the similarity and the cophenetic matrices was 0.968. The use of the Jaccard similarity coefficient or the change of the order in which isolate data were input into the software did not lead to significant changes in the final dendrogram. The UPGMA method was chosen because it has been shown to give the most accurate representation of data (18).

View larger version (98K): [in this window] [in a new window]   FIG. 1.   Example of the ribopatterns obtained with the B. cepacia isolates indicated below after digestion of total DNA with EcoRI and hybridization with the acetylaminofluorene-labeled 16S + 23S rRNA from E. coli. Lane 1, molecular size standard RaoulI; lanes 2 to 12, B. cepacia isolates IST403, IST406, IST407, IST401, IST408, IST410, IST404, IST411, IST409, IST402, and ATCC25416, respectively. Sizes in kilobases are given on the left.

View larger version (18K): [in this window] [in a new window]   FIG. 2.   Dendrogram showing the results of clustering analysis using UPGMA for the B. cepacia strains and isolates under study. The numbers in the horizontal axis indicate the percentage of similarity as determined with the Dice coefficient for the ribopatterns schematically represented on the left, with molecular sizes (in kilobases) arranged in a logarithmic scale.

View larger version (55K): [in this window] [in a new window]   FIG. 3.   Comparison of the AseI macrorestriction fragment patterns of the genomic DNAs from B. cepacia isolates separated by PFGE. (A) Lanes: 1, size standard of concatamerized phage  DNA; 2, IST409; 3, IST412; 4, IST413; 5, IST414; 6, IST415; 7, IST402. (B) Lanes: 1, IST401; 2, IST406; 3, IST408; 4, size standard of concatamerized phage  DNA.

EPS biosynthesis. The mucoidy of B. cepacia isolates was assessed by comparing the morphologies of isolated colonies formed after incubation for 5 days at 30°C in the three different media examined, supplemented with 20 g of agar (Iberagar; Coina, Portugal) per liter. CDM and A media were described in the literature as leading to the production of EPS by a specific B. cepacia CF isolate (1, 20), and S medium was successfully used in the IST laboratory to overproduce the EPS gellan gum from Sphingomonas paucimobilis ATCC 31461 (8, 15, 17). Moreover, the use of S agar plates allowed the differentiation of the mucoid colonial morphotypes of S. paucimobilis variants, which produced mutated gellan gum in different yields (15, 17). CDM contained, in grams per liter, NaCl (0.175), KCl (0.224), (NH₄)₂SO₄ (0.396), K₂HPO₄ (0.205), and glucose (10). Medium A contained, in grams per liter, only yeast extract (2) and glucose (10). S medium contained, in grams per liter, Na₂HPO₄ (10), KH₂PO₄ (3), K₂SO₄ (1), NaCl (1), MgSO₄ · 7H₂O (0.2), yeast extract (Difco) (1), Casamino Acids (Difco) (1), CaCl₂ · 2H₂O (0.01), FeSO₄ · 7H₂O (0.001), and glucose (20). Isolated colonies were obtained by spreading, after suitable dilution, 100 µl of liquid culture onto the surface of the solid growth medium. The liquid culture used to inoculate the agar plates resulted from overnight cultivation at 30°C with orbital agitation (250 revolutions · min¹) in Luria broth (Sigma) medium. EPS production by the different isolates was quantified after 5 days of incubation at 30°C of confluent cell growth in either A or S solid medium. The plates were scraped, the material was resuspended in 0.9% (wt/vol) NaCl by vortexing, and the cells were separated by centrifugation at 20,000 × g for 15 min. The EPS was precipitated from the cell-free supernatant by the addition of 2 volumes of cold ethanol, air dried, and redissolved in distilled water. The total sugar content was assessed by the phenol-sulfuric acid method (7) using the EPS produced by isolate IST408 as a standard. For this purpose, the EPS produced by isolate IST408 was further dialyzed against distilled water at 4°C for 24 h and recovered by freeze-drying. The cell pellets obtained from each plate were washed once with 0.9% NaCl, and the protein content was quantified by the biuret method (12) using bovine serum albumin fraction V (Merck) as a standard. The results of EPS production were expressed as grams of total sugars per gram of protein and are the means of at least three independent cultivations and of three determinations of total sugar and protein contents in each independent sample. After 5 days of incubation at 30°C on solid CDM medium, no mucoid colonies were detected by visual inspection. Confirming this observation, we were unable to detect any ethanol-precipitable material from the cell-free supernatants. However, when grown on solid medium A, isolates IST401, IST406, and IST408 formed colonies that evidenced a very clear mucoid phenotype, while the other isolates maintained the nonmucoid appearance (results not shown). The indications of this visual observation were confirmed by quantitative analysis of EPS production (Table 1). The use of solid S medium to cultivate the whole collection of isolates allowed the identification of 12 Portuguese isolates (out of 16) giving rise to mucoid colonies after 5 days of incubation at 30°C. The quantification of the EPS produced by the isolates during growth in this solid medium confirmed the colony morphotypes (Table 1). Only the isolates IST405, IST407, IST410, and IST411 (4 out of 16) were consistently unable to produce EPS in any of the three media tested. Among the three CF epidemic isolates used as reference strains, only one was able to produce EPS in S medium while the environmental type strain ATCC 25416 produced high levels of EPS (Table 1). In general, the relative level of EPS produced on S plates by the different isolates was reproduced in S liquid culture carried out in shake flasks (results not shown). The structural analysis of the EPS produced by IST408 in A agar plates indicated that the polymer is very similar, if not identical, to the EPS produced by a French clinical isolate (5; P. Cescutti, M. Bosco, F. Picotti, J. A. Richau, J. H. Leitão, and I. Sá-Correia, Abstr. 10th Eur. Carbohydr. Symp., abstr. OB05, p. 85, 1999), and a detailed description will be published elsewhere. Moreover, the sugar composition of the EPS from B. cepacia IST408 (composed of glucose, mannose, rhamnose, galactose, and glucuronic acid in the molar ratio 1.0:1.0:1.0:3.0:1.0 [Cescutti et al., Abstr. 10th Eur. Carbohydr. Symp.]) is similar to the compositions of the EPSs produced by B. cepacia isolates from patients with CF in the United States (20) and in the United Kingdom (1).