Identification of Isolates within the Burkholderia cepacia Complex by a Multiplex recA and 16S rRNA Gene Real-Time PCR Assay

Although identification of members of the Burkholderia cepacia complex (BCC) can be accomplished with phenotypic methods (5), this tends to remain the domain of reference laboratories. However, the genetic sequence of the recA gene has also been found to be useful for BCC genomovar identification (6, 8). We describe the use of a monochrome (SYBR green I), real-time PCR assay with simultaneous detection of the 16S rRNA and recA genes followed by sequenced-based identification.

This method was evaluated using stored DNA from 99 isolates (collected from 48 patients) previously referred to our molecular diagnostic laboratory for BCC culture confirmation by conventional PCR (8), from 2001 through 2006. A method that we previously described for real-time partial 16S rRNA gene PCR (10) was modified by the inclusion of recA gene primers (BCR1/BCR2, 10 pmol) (8) and alteration of the extension time to 40 seconds. Melting curve analysis was used to detect the presence of the 1,040-bp recA gene amplicon and to differentiate it from the 500-bp partial 16S rRNA gene amplicon. Sequencing of the respective amplicons was performed as previously described (10).

A BLAST search (1) was used to assign a sequence-based identification to each organism. Although the 16S rRNA gene sequence cannot be used to differentiate the members of the BCC (2), we found that a BLAST match of the recA gene sequence provided clear distinction between genomovars for our 51 BCC-positive isolates. Distinction of Burkholderia cenocepacia lineages III-A and III-B was also possible after examination of the papers in which the matching sequences were published (8, 9, 11). When the recA gene sequences of our isolates were aligned, the differences in nucleotide similarity confirmed this distinction. For this evaluation, we also confirmed the identity of each isolate by using genomovar-specific primers (3, 8).

The predominant genomovar among our BCC isolates was B. cenocepacia III-B (26 isolates from 13 patients), followed by Burkholderia multivorans (20 isolates from 9 patients). Two isolates of B. cenocepacia III-A were identified from a patient from the United Kingdom, and another was isolated from a patient from North America. One isolate each of Burkholderia ambifaria and Burkholderia vietnamiensis was identified as well. The GenBank accession numbers for the recA gene sequences of these isolates are EF427789 to EF427839.

In addition to recA gene-based BCC identification, 16S rRNA gene sequencing permitted the identification of the non-BCC gram-negative isolates. These isolates included Achromobacter xylosoxidans, Burkholderia gladioli, Pandoraea sputorum, Pseudomonas aeruginosa, Pseudomonas putida, Ralstonia mannitolytica, Ralstonia pickettii, Sphingomonas paucimobilis, and Stenotrophomonas maltophilia. These organisms can be difficult to identify (7) and may, at times, be clinically relevant (4). The GenBank accession numbers for the 16S rRNA gene sequences of all the isolates are EF427693 to EF427788.

It is important to remember, however, that phenotypic characters and sequence database integrity remain integral to sequence-based identification. We also acknowledge that cblA gene PCR needs to be employed to distinguish the ET12 strain of B. cenocepacia III-A (12). Furthermore, we note that referral to a reference laboratory is still necessary for any formal strain typing that may be required.

In summary, we consider that this method provides accurate identification of BCC genomovars from culture. Moreover, we believe that this level of BCC identification is both practical and appropriate for a hospital laboratory to provide.