HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via CrossRef Citing Articles via Google Scholar Google Scholar Articles by VERMIS, K. Articles by VANDAMME, P. Search for Related Content PubMed PubMed Citation Articles by VERMIS, K. Articles by VANDAMME, P. Agricola Articles by VERMIS, K. Articles by VANDAMME, P.

Evaluation of species-specific recA-based PCR tests for genomovar level identification within the Burkholderia cepacia complex

KAREN VERMIS, TOM COENYE^*, ESHWAR MAHENTHIRALINGAM, HANS J. NELIS and PETER VANDAMME^*

Laboratory of Pharmaceutical Microbiology and *Laboratory of Microbiology, Ghent University, Ghent, Belgium and Cardiff School of Biosciences, Cardiff University, Cardiff

Corresponding author: Dr P. Vandamme (e-mail: peter.vandammme{at}rug.ac.be).

Received 2 April 2002; accepted 14 June 2002.

Abstract

The Burkholderia cepacia complex presently comprises nine genomovars: B. cepacia (genomovar I), B. multivorans (genomovar II), B. cepacia genomovar III, B. stabilis (genomovar IV), B. vietnamiensis (genomovar V), B. cepacia genomovar VI, B. ambifaria (genomovar VII), B. anthina (genomovar VIII) and B. pyrrocinia (genomovar IX). Strains of each genomovar can colonise the respiratory tract of cystic fibrosis (CF) patients. However, the majority of infections in CF patients are caused by B. multivorans and B. cepacia genomovar III isolates. Accurate genomovar-level identification is best achieved through a polyphasic approach combining phenotypic and genotypic analyses. In the present study, the sensitivity and specificity of recA-based genomovar specific primer pairs were evaluated with a collection of 508 B. cepacia complex isolates representing all nine genomovars. The assays for the identification of B. multivorans (sensitivity and specificity, 100%), B. cepacia genomovar III (sensitivity, 92%; specificity, 100%), and B. ambifaria (sensitivity and specificity, 100%) were the most efficient. However, the B. cepacia genomovar I assay lacked sensitivity (72%) and cross-reacted with all B. pyrrocinia isolates examined. Several new recA RFLP types were also revealed within the B. cepacia complex. One of these profiles was shared by a clinical and an environmental B. cepacia-like isolate and by the B. ubonensis type strain. The latter organism is a recently described soil bacterium. Its relationship to the various B. cepacia complex genomovars needs further study.

The clinical significance of Burkholderia cepacia, originally described as a plant pathogen in the 1950s [1], increased in the 1980s as it emerged as an opportunist pathogen, colonising the lungs of cystic fibrosis (CF) patients [2–4] and immunocompromised individuals [5–7]. At its peak incidence in the mid-1980s, c. 20% of colonised patients suffered from severe progressive respiratory failure with bacteraemia (`cepacia syndrome') which was often fatal [2, 8]. Accumulated evidence of nosocomial acquisition and person-to-person transmission led to the implementation of stringent infection control measures, which resulted in a decline of new infections [9–11]. However, it was apparent that not all B. cepacia strains have the same capacity for epidemic cross-infection or are capable of causing fatal pneumonia [8, 12, 13]. Taxonomic studies have revealed that B. cepacia consists of a group of at least five phenotypically similar and closely related genomic species (B. cepacia genomovars I–V), which has become known as the B. cepacia complex [14]. Recently, four further taxa (B. cepacia genomovars VI–IX) have been described as members of the complex [15–17]. Discrimination between several of these genomovars is sometimes difficult. A binomial name was assigned to those genomovars that could be differentiated phenotypically, i.e., B. multivorans (formerly genomovar II) [14], B. stabilis (genomovar IV) [18], B. ambifaria (genomovar VII) [16] and B. anthina (genomovar VIII) [17]. Genomovar V and genomovar IX have been recognised as the previously described species B. vietnamiensis [19] and B. pyrrocinia, respectively [20]. B. cepacia genomovar III and B. cepacia genomovar VI cannot be differentiated in biochemical tests from B. cepacia genomovar I and B. multivorans, respectively.

Currently, in our laboratories accurate identification of members of the B. cepacia complex is the result of a polyphasic approach, including biochemical tests, whole-cell protein electrophoresis, DNA–DNA hybridisation and recA restriction fragment length polymorphism (RFLP) analysis. A single test procedure to distinguish between all genomovars is not yet available. Molecular diagnostic tests based on species-specific rDNA PCR [21–23] and PCR-RFLP of 16S rDNA [24, 25] are useful to identify B. cepacia genomovars II, IV, V and VI. However, B. cepacia genomovars I, III and VII cannot be distinguished from each other. Moreover, there are no published data for B. anthina and B. pyrrocinia [24]. In recent studies, Mahenthiralingam and co-workers reported that recA gene sequence comparison offered a superior potential for the development of genomovar-specific PCR tests [16, 26]. At present, specific primers are available for the detection of B. cepacia genomovar I, B. multivorans, B. cepacia genomovars III-A and III-B (two distinct recA-based phylogenetic subgroups within genomovar III), B. stabilis, B. vietnamiensis [26] and B. ambifaria ([16]. An error in one of the published primer sequences was corrected in an erratum available at http://ijs.sgmjournals.org/cgi/content/full/51/4/1481/DC1.

Although all B. cepacia genomovars can colonise CF patients, the large majority of infections is caused by B. multivorans or B. cepacia genomovar III [12, 13, 27]. The aim of the present study was to evaluate the sensitivity and specificity of recA-based PCR tests for the identification of these B. cepacia genomovars and of PCR tests for B. cepacia genomovar I and B. ambifaria. The latter genomovars are very difficult to distinguish from B. cepacia genomovar III. Whereas the original set of isolates used to design the recA-based PCR tests and to evaluate their sensitivity and specificity was primarily of CF origin, the 508 isolates of the present study were obtained from a range of environmental sources and from human infections (CF and non-CF). They were isolated in 19 countries in Europe, North and South America, Asia and Australia.

Materials and methods

Bacterial strains

All isolates were grown aerobically at 28°C on Trypticase Soy Agar (Becton Dickinson, Sparks, MD, USA). Species or genomovar assignment was based on a combination of results obtained from whole-cell protein electrophoresis [14] and recA RFLP analysis [26]. B. cepacia genomovar I was represented by 29 clinical and 42 environmental isolates, B. multivorans by 85 clinical and 17 environmental isolates, B. cepacia genomovar III by 172 clinical and 62 environmental isolates, B. stabilis and B. vietnamiensis by 6 clinical and 1 environmental isolate each, B. cepacia genomovar VI by 4 clinical isolates, B. ambifaria by 6 clinical and 47 environmental isolates, B. anthina by 5 clinical and 15 environmental isolates and B. pyrrocinia by 7 environmental isolates. Two isolates (1 clinical and 1 environmental) with a recA RFLP pattern different from the recognised patterns within the B. cepacia complex were also included. The recA RFLP profile of the latter isolates was identical to that of B. ubonensis LMG 20358^T, a recently described soil bacterium [28], which was also included in the present study.

DNA preparation and PCR assays

DNA was prepared as described before [16]. PCR assays for the identification of B. cepacia genomovars I and III, B. multivorans and B. ambifaria were performed as described previously [16, 26] with the following modifications. The annealing temperature for the PCR tests for B. multivorans and B. cepacia genomovars III-A and III-B was raised to 64°C, 68°C and 68°C, respectively, to eliminate non-specific primer binding. Furthermore, the correct primer sequence for primer BCRGC2 used for the identification of B. ambifaria is 5'-TCC GCA GCC GCA CCT TCA-3'. The PCR apparatus used was a GeneAmp PCR System 9600 (Perkin-Elmer Applied Biosystems, Foster City, CA, USA).

recA gene sequencing and RFLP analyses

recA gene sequencing and RFLP analyses were performed as described before [26]. The nucleotide sequence of the entire recA gene from the following strains was determined and deposited in GenBank (GenBank accession no. shown in brackets): LMG 14095 (AF456016), LMG 6863 (AF456019), LMG 18943 (AF456029), L06 (AF456030), LMG 14294 (AF456031), ATCC 39277 (AF456032), LMG 6964 (AF456065), LMG 6963 (AF456068), R-10741 (AF456084) and R-14268 (AF456122). These sequences were aligned by use of the CLUSTAL W algorithm with recA sequences from B. cepacia complex strains analysed in previous studies and a neighbour-joining genetic distance phylogenetic tree constructed as described previously [16, 18, 26].

DNA–DNA hybridisation

The methodology used for DNA–DNA hybridisation experiments (including DNA preparation) was as described previously [15]. The hybridisation temperature was 50°C.

Results and discussion

Each positive PCR test resulted in an amplification product of the expected size [16, 26].

Fifty-one of 71 isolates of B. cepacia genomovar I gave positive results with the genomovar I specific primers (sensitivity, 71.8%). No cross-reactions were observed with primers specific for other genomovars. All but one of the 20 remaining B. cepacia genomovar I isolates belonged to two new recA RFLP profiles within B. cepacia genomovar I. RecA sequence analysis confirmed that one of these types, designated AW, clusters close to previously characterised genomovar I recA sequences [26] (Fig. 1). The second type, designated K, occupied a distinct position in the phylogenetic tree. DNA–DNA hybridisation between two of these recA type K isolates (LMG 14095 and LMG 6863) and the B. cepacia genomovar I type strain (LMG 1222) revealed hybridisation values of 71% and 75%, respectively, confirming their identity as B. cepacia genomovar I.

View larger version (23K): [in this window] [in a new window]   Fig. 1. Phylogenetic tree of recA sequences from strains of the current B. cepacia complex species. The tree was constructed as described in Materials and methods. Genetic distance is shown on the scale and bootstrap analysis for node values >70% are indicated. The B. cepacia complex genomovar or species, and the positions of strains of novel recA RFLP type AD, AW and K are indicated.

Ninety-seven of 234 B. cepacia genomovar III isolates gave positive results in the B. cepacia genomovar III-A PCR test, and 114 B. cepacia genomovar III isolates reacted with the B. cepacia genomovar III-B primer pair (combined sensitivity for B. cepacia genomovar III, 92.2%). Five B. cepacia genomovar III isolates generated an amplicon with both the III-A and III-B specific primer pairs and 18 genomovar III strains failed to react with either of the genomovar III specific primer pairs. The recA RFLP patterns of these false-negative results were heterogeneous. Six strains were of recA RFLP type H, five strains belonged to type I, four strains to G and three strains to a novel RFLP type, designated AD. Phylogenetic analysis confirmed that type AD isolates clustered closely with B. cepacia genomovar III-B strains (Fig. 1). The B. cepacia genomovar III isolates did not react with primers specific for other genomovars.

All B. multivorans and B. ambifaria isolates examined generated an amplicon with their respective primer pair (100% sensitivity). No cross-reactions were observed with primer pairs specific for other genomovars.

None of the B. stabilis, B. vietnamiensis, B. cepacia genomovar VI and B. anthina isolates reacted with the primer pairs studied. However, all seven B. pyrrocinia isolates gave positive results in the B. cepacia genomovar I specific PCR test; no amplification products were generated in the other PCR assays.

The three isolates (with identical RFLP profiles) which were not assigned to one of the B. cepacia complex genomovars (B. ubonensis LMG 20358^T and two B. cepacia-like strains) gave negative results in all of the PCR tests examined. The fact that these isolates generated a recA amplicon with the B. cepacia complex specific primers suggested that they are related to the complex. These findings confirmed the high level of 16S rDNA similarity reported between B. ubonensis and other B. cepacia complex bacteria [28]. Further taxonomic work is required to clarify the status of these isolates within the B. cepacia complex.

In summary, the recA-based PCR tests examined proved to be excellent diagnostic tools for the identification of B. cepacia complex members. They allowed identification of the whole B. cepacia complex as a group and of B. multivorans and B. cepacia genomovar III, the prominent organisms in CF specimens. However, the sensitivity and specificity of the B. cepacia genomovar I assay should be improved, particularly for environmental samples. The diverse strain collection examined in the present study revealed several new recA RFLP types within the B. cepacia complex and included one CF and one environmental isolate that were indistinguishable from B. ubonensis, another putative member of the B. cepacia complex.

Acknowledgments

We are indebted to the Fund for Scientific Research - Flanders (P.V.) and the UK Cystic Fibrosis Trust (E.M. and P.V) for financial support. E.M. is grateful to Julie Fadden for excellent technical assistance.

References