OUP user menu

Conservation of a novel protein associated with an antibiotic efflux operon in Burkholderia cenocepacia

Bindu M. Nair, Lukasz A. Joachimiak, Sujay Chattopadhyay, Idalia Montano, Jane L. Burns
DOI: http://dx.doi.org/10.1016/j.femsle.2005.03.027 337-344 First published online: 1 April 2005


Burkholderia cenocepacia is a significant problem in individuals with cystic fibrosis and is a member of the B. cepacia complex of closely related antibiotic resistant bacteria. A salicylate-regulated antibiotic efflux operon has been identified in B. cenocepacia and one of its four genes, llpE, is without parallel in previously reported efflux operons. PCR amplification and sequencing of llpE from B. cepacia complex isolates demonstrated the highest prevalence in B. cenocepacia with a high degree of sequence conservation. While at least one non-synonymous mutation was identified between isolates from different genomovars, only synonymous differences were identified within the IIIA and IIIB sub-groups of B. cenocepacia. Structural modeling suggests that LlpE is a member of the α/β hydrolase enzyme family. Identification of strong structural homology to hydrolases and a high degree of conservation in B. cenocepacia suggests an enzymatic function for LlpE, benefiting survival in the cystic fibrosis lung.

  • Burkholderia cenocepacia
  • Antibiotic efflux
  • Hydrolytic enzymes

1 Introduction

Burkholderia cepacia complex strains are important pulmonary pathogens in patients with cystic fibrosis. Although isolated from only 3.1% of patients over the age of five [[], this organism may be associated with severe morbidity and mortality in the cystic fibrosis population [[]. The complex is a group of at least ten related genomovars [[[]. Although virtually all genomovars have been isolated from individuals with cystic fibrosis, most cystic fibrosis isolates are genomovars II and III, now called B. multivorans and B. cenocepacia, respectively [[,[]. Of these, B. cenocepacia has been most commonly associated with epidemic spread and increased clinical virulence [[,[0]. Based on recA typing, B. cenocepacia has subsequently been divided into four subgroups, of which genomovars IIIA and IIIB are most prevalent in cystic fibrosis [[].

Multiple antibiotic resistance is a characteristic of all B. cepacia complex isolates, particularly those from individuals with cystic fibrosis. Antibiotic efflux is a common mechanism of resistance in Gram-negative bacteria [[1]. Efflux pumps have been implicated in resistance to diverse antibiotics [[2[9] and other toxic compounds including dyes, detergents, disinfectants, and fatty acids [[1,[9].

We previously identified efflux-mediated multidrug resistance to chloramphenicol, trimethoprim, and ciprofloxacin in B. cenocepacia[[0,[1]. The ceo efflux operon responds to salicylate as an inducer and as a substrate of efflux [[1]. The genes ceoA, ceoB, and opcM have significant homologies to the resistance/nodulation/cell division family of proteins, members of which are components of multidrug efflux systems. A fourth gene, llpE, has been identified in the cluster and is co-transcribed with ceoA, ceoB, and opcM[[1], but its role in efflux is unclear. There is no counterpart of LlpE in any other efflux operons characterized to date. In this study, we elaborate on the potential role of LlpE in efflux, examining its prevalence in the B. cepacia complex, establishing a putative structural model for LlpE based on homologous proteins, and speculating on a possible enzymatic function in cystic fibrosis.

2 Materials and methods

2.1 Culture conditions

Strains used in this study are listed in Table 1. B. cepacia complex strains were grown in L-broth at 37 °C, 250 rpm.

View this table:
Table 1

Strains used in this study

K61-3B. cenocepacia, IIIA, CF clinical isolate, original source for ceo efflux operon[[0]
ATCC 25416B. cepacia, onion isolate[[6]
ATCC 17759B. cepacia, soil isolate[[6]
CEP509B. cepacia, CF isolate[[6]
LMG 17997B. cepacia, UTI isolate[[6]
C5393B. multivorans, CF isolate[[6]
LMG 13010B. multivorans, CF isolate[[6]
C1576B. multivorans, CF isolate[[6]
CP-A1-1B. multivorans, CF isolate[[6]
JTCB. multivorans, CGD isolate[[6]
C1962B. multivorans, abscess isolate[[6]
249-2B. multivorans, laboratory isolate[[6]
ATCC 17616B. multivorans, soil isolate[[6]
J2315B. cenocepacia, IIIA, CF isolate[[6]
BC7B. cenocepacia, IIIA, CF isolate[[6]
K56-2B. cenocepacia, IIIA, CF isolate[[6]
C5424B. cenocepacia, IIIA, CF isolate[[6]
C6433B. cenocepacia, IIIA, CF isolate[[6]
C1394B. cenocepacia, IIIB, CF isolate[[6]
PC184B. cenocepacia, IIIB, CF isolate[[6]
CEP511B. cenocepacia, IIIB, CF isolate[[6]
J415B. cenocepacia, IIIB, CF isolate[[6]
ATCC 17765B. cenocepacia, IIIB, UTI isolate[[6]
LMG 14294B. stabilis, CF isolate[[6]
C7322B. stabilis, CF isolate[[6]
LMG 18888B. stabilis, non-CF clinical isolate[[6]
LMG 14086B. stabilis, ventilator isolate[[6]
PC259B. vietnamiensis, CF isolate[[6]
LMG 16232B. vietnamiensis, CF isolate[[6]
FC441B. vietnamiensis, CGD isolate[[6]
LMG 10929B. vietnamiensis, rice isolate[[6]
CEP021B. dolosa, CF isolate[[7]
FC353B. dolosa, CF isolateJ. LiPuma
ATCC53266B. ambifaria, soil isolate[[7]
FC876B. ambifaria, soil isolateT. Heulin
ATCC17760B. ambifaria, soil isolateATCC
  • a Abbreviations: CF, cystic fibrosis; UTI, urinary tract infection; CGD, chronic granulomatous disease.

2.2 PCR conditions

Genomic DNA for amplification of llpE was isolated from B. cepacia complex isolates using the DNeasy tissue Kit (Qiagen, Valencia, CA). For assessing the prevalence of llpE, genomic DNA was PCR amplified using primers llpE-F1 (5′GGCCTGGAAGCTTGCTTCGG3′) and llpE-R (5′GCATTAGTCCATGGTTATTCGGGACGGTTCGGC 3′) [[1]. PCR used the GC Rich PCR system (Roche Applied Science, Indianapolis, IN), 100 ng of template DNA, 20 pmol of each primer, and 200 μM deoxynucleotides. An initial denaturation at 95 °C for 4 min was followed by 30 cycles of denaturation at 95 °C for 30 s, annealing at 66 °C for 30 s and elongation at 72 °C for 60 s, with a 5 s increment in elongation temperature per cycle for the last 20 cycles. After a final extension at 72 °C for 7 min, the samples were examined by agarose gel electrophoresis. PCR bands from selected strains were gel-purified and sequenced using the ABI PRISM BigDye Terminator Cycle Sequencing Ready Reaction Kit (Applied Biosystems/Perkin–Elmer, Foster City, CA).

2.3 Phylogenetic analysis

Based on zonal analysis, a recent tool for visualizing gene evolution [[2], an unrooted protein phylogram was constructed based on llpE coding sequences from 12 strains of the B. cepacia complex. This phylogram was created to give information on inter- and intra-nodal synonymous variations, where each node represents a structural variant. ClustalX 1.83 was used to align the various DNA and protein sequences. The aligned DNA dataset was the input in the PAUP∗4.0b program to generate a maximum likelihood unrooted DNA phylogram, which was finally converted to a protein phylogram incorporating the synonymous and non-synonymous changes from the aligned protein sequence dataset.

2.4 Homology modeling

Comparative models for LlpE were built with the help of the Robetta server [[3,[4]. Robetta is a fully automated full-chain protein structure prediction server and the LlpE model was built from a homologous protein (E-value 2E-28, sequence identity 38%) detected with BLAST [[5] and aligned by the K∗Sync alignment method [[3]. Loop regions were assembled ab initio by the Rosetta program fragments and optimized to fit the aligned template structure [[6].

3 Results

3.1 Sequence analysis

The region upstream of ceoABopcM from K61-3 was sequenced and shown to code for a 286 amino acid protein. This gene, termed llpE, is the first gene in the ceo antibiotic efflux operon [[1]. The N-terminal 70 amino acids of LlpE were analyzed for presence of signal peptide, using SignalP V2.0 (http://www.cbs.dtu.dk/services/SignalP-2.0/) [[7]. LlpE is predicted to be a non-secretory protein and to have a signal peptide that is cleaved between amino acids 22 and 23. Analysis of LlpE by PSORT (http://psort.nibb.ac.jp/) [[8], a program for prediction of protein localization, predicts a periplasmic or outer membrane localization.

3.2 Identification of homologs

Subjecting LlpE to a BLAST search [[5] did not reveal homology to previously characterized B. cepacia lipases, none of which are associated with antibiotic resistance [[9,[0]. However, the search did reveal homology to various eukaryotic and prokaryotic lipases and esterases belonging to the α/β hydrolase family of proteins. These proteins have diverse functions, but share a common three-dimensional structure, including α/β hydrolase folds, often in the absence of detectable homology [[1[4]. Multiple sequence alignment of nine of the most homologous proteins using the BCM Search Launcher [[5] and subsequent BOXSHADE analysis reveals alignment of a consensus sequence Sm-X-Nu-X-Sm-Sm, where Sm is a small amino acid like glycine, X is any amino acid and Nu is the nucleophilic residue (G-H-D132-A-G-G in LlpE). This sequence is conserved in the GXD/SXG family of lipolytic enzymes and the Type B family of carboxylesterases. The most homologous proteins included: 1JJIA, a novel hyper-thermophilic carboxylesterase from the archaeon A. fulgidus; AaaD, human arylacetamide deacetylase; Aes, an Escherichia coli acetyl esterase, Est, an esterase from Acinetobacter lwoffii; Bah, an acetyl hydrolase from Streptomyces hygroscopicus; LipS, a human hormone sensitive lipase (HSL); VSH5, Dictyostelium discoideum vegetative specific protein H5; Est-P, Drosophila melanogaster esterase P precursor; and Est-1, a human liver carboxylesterase 1 precursor. All of these proteins contain a catalytic triad in the conserved order of nucleophilic residue-acidic residue-histidine, which in LlpE is predicted to be D132, D226 and H256. In all the solved structures of these proteins, the triad is accommodated on loops that are the best-conserved features of these proteins.

3.3 Prevalence among genomovars

Genomic DNA was harvested from 37 strains from seven genomovars belonging to the B. cepacia complex [[6,[7] and subjected to PCR amplification for llpE. PCR conditions were optimized using genomic DNA from B. cenocepacia K61-3, to show amplification of a single 630 bp band. Agarose gel analysis revealed a single 630 bp band for all five B. cepacia (genomovar I) strains, all 11 B. cenocepacia strains, and four of five B. stabilis (genomovar IV) strains (Fig. 1). Multiple bands or a single band of the wrong size were amplified from all strains belonging to B. multivorans, B. vietnamiensis, B. dolosa and B. ambifaria, as well as from the remaining B. stabilis strain.

Figure 1

PCR amplification of llpE from B. cepacia complex experimental panel strains. III: B. cenocepacia, I: B. cepacia, II: B. multivorans, IV: B. stabilis, V: B. vietnamiensis, VI: B. dolosa, VII: B. ambifaria, mw: molecular weight markers (identical in all gels) with sizes shown in Kb. Lanes 1: J2315, 2: C5424, 3: PC184, 4: J415, 5: BC7, 6: C6433, 7: CEP511, 8: ATCC 17765, 9: K56-2, 10: C1394, 11: K61-3 (source of cloned llpE), 12: ATCC 25416, 13: ATCC 17759, 14: CEP509, 15: LMG 17997, 16: K61-3, 17: C5393, 18: LMG 13010, 19: C1576, 20: CP-A1-1, 21: JTC, 22: C1962, 23: 249-2, 24: ATCC 17616, 25: LMG 14294, 26: C7322, 27: LMG 18888, 28: LMG 14086, 29: K61-3, 30: PC259, 31: LMG 16232, 32: FC441, 33: LMG 10929, 34: CEP21, 35: FC353, 36: ATCC17760, 37: FC876, 38: ATCC53266.

3.4 Phylogenetic analysis

The 630 bp PCR bands from all 11 B. cenocepacia strains and one representative strain each from B. cepacia and B. stabilis were gel-purified and sequenced to examine genetic relatedness. In Fig. 2, each node is represented by a single genomovar group (example, genomovar I and IV) or sub-group (example, IIIA and IIIB). Phylogenetic analysis of B. cepacia and B. stabilis revealed 19 variations in the 2 llpE gene sequences, of which 15 are synonymous and 4 are non-synonymous. B. cenocepacia isolates, including distinct IIIA and IIIB subgroups, were also separated from B. cepacia and B. stabilis and from each other by synonymous and non-synonymous mutations However, within these two subgroups there were no non-synonymous mutations. Analysis of genomovar IIIB reveals that the node consists of five strains differentiated by 21 changes in the llpE gene, all of which are synonymous, whereas in genomovar IIIA, strain C6433 differs from the five other genomovar IIIA strains by two synonymous mutations. Based on the extent of sequence differences, ATCC25416 (genomovar I) and LMG14294 (genomovar IV) appear as outgroups compared to genomovar III subgroups. Sequence analysis of selected non-630 bp PCR bands did not show homology to llpE (data not shown).

Figure 2

Zonal analysis diagram. From the unrooted maximum likelihood tree, the alleles differing only in synonymous variations are collapsed into a single node. The branch(es) connecting any two nodes corresponds to the number of non-synonymous changes and the amino acid changes and positions are shown along the branches in italics. The values along the branches represent the numbers of inter-nodal synonymous mutations between the two respective nodes. The number of alleles in each of the two multiple-allele nodes is shown along with the number of intra-nodal synonymous changes in parentheses. Since each node (i.e., each structural variant) represents a particular genomovar group or subgroup, the genomovars are denoted beside the nodes.

3.5 Putative model of LlpE

Conservation of the catalytic triad has been corroborated in LlpE by homology modeling of the LlpE structure using the Robetta server [[3,[4]. The template for the LlpE model was detected by BLAST and derived from the structure of a novel hyper-thermophilic carboxylesterase from the archaeon, Archaeoglobus fulgidus (Protein data bank ID:1JJI) [[8]. This protein demonstrates a canonical α/β hydrolase core that contains a central-sheet composed of eight strands. The core is shielded on the C-terminal side by a cap region constituted of two separate regions (amino acid residues 1–54 and 188–246) at the carboxyl edge of the central β-sheet. The putative LlpE model has a near identical core region to the parent structure, but has a simplified cap (Fig. 3(a)). The cap region is comprised of the equivalent region on the parent structure (188–246), with a truncated and unstructured N-terminal portion, which is consistent with it being a signal sequence for export into the periplasm [[1]. Residues in the putative catalytic triad in LlpE (D132, D226, and H256) occupy canonical positions, where the nucleophile aspartic acid (132) is located within a sharp turn known as the nucleophilic elbow (Fig. 3 (b)). At its apex, this turn contains the signature motif GXD/SXG, which characterizes all members of this family of proteins. Analysis using JEvTrace [[9] revealed that the highly conserved GXD/SXG motif fits nicely in a pocket at the base of the β-sheet and leads to a large ovoidal tunnel that could accommodate a large substrate with approximate dimensions of 10 Å× 12 Å (Fig. 3 (b)). It is conceivable that the almost planar substrates of the ceo efflux system including the antibiotics chloramphenicol, trimethoprim, and ciprofloxacin, and the siderophore, salicylate, with dimensions of 9 × 3 × 4 Å, 9 × 7 × 4 Å, 11 × 6 × 1 Å, and 4 × 3 Å, respectively, can be accommodated in this tunnel during efflux.

Figure 3

Putative model of LlpE. (a) Topology map of the LlpE model generated by the Robetta server. The N-terminal cap region is truncated relative to the parent 1JJI but the remainder of the core structure is nearly identical to the Afest fold. (b) Surface representation of the LlpE model in teal; in orange are the nucleophilic elbow residues(130–134, including D132); in yellow are the active site residues D132, D226 and H256; in red are external loops R175–I180, D164–R169 and L193–Q195; and in white are the internal loops H58–V63 and G160–L163. At the bottom is a zoom of the active site residues, colored as before.

The close structural and sequence similarity of the LlpE model to the parent structures allows definition of equivalent regions. As in the A. fulgidus structure, the loop regions R175–I180, D164–R169 and L193–Q195 define the entrance to the tunnel whereas the loop regions H58–V63 and G160–L163 define the internal borders. Further, the homologous crystal structure revealed the presence of a covalent adduct to a piperazine-ethane moiety at the active site S160, which has been speculated to be the acyl-binding pocket [[8]. The location of the two other catalytic residues, D226 and H256, coordinate the proximal function of these residues as a charge relay network and proton carrier, respectively.

4 Discussion

Originally known as B. cepacia genomovar III [[], B. cenocepacia can be a life-threatening multidrug resistant pathogen and is the most commonly encountered genomovar in patients with cystic fibrosis [[,[]. We previously demonstrated salicylate (and iron) regulated antibiotic efflux in B. cenocepacia strain K61-3 [[0,[1]. The current study investigated a novel protein, LlpE, that is a component of the ceo antibiotic efflux operon. Prevalence of the gene within the B. cepacia complex and phylogenetic relatedness within and among genomovars was examined. A structural model was generated based on sequence homologies with other α/β hydrolase family proteins.

Of the seven genomovars examined, llpE homologs were found only in representatives of genomovars I, III, IV and VI. All B. cepacia and B. cenocepacia isolates tested by PCR amplification, demonstrated llpE. Subjecting llpE sequences from these strains to phylogenetic analysis, revealed the presence of at least one non-synonymous variation between genomovar groups I, IIIA, IIIB, and VI but none within individual B. cenocepacia subgroups. LlpE from genomovars IIIA and IIIB differ in the amino acid residue at LlpE position 234. An asparagine (N) occupies this position in genomovar IIIA strains, while a lysine (K) occupies it in genomovar IIIB strains. Based on the LlpE model, N234 is in helix 8 immediately after the putative active residue D226. The mutation N234K (in genomovar IIIB) might result in salt bridges between K234 and either D164 or E208. It is also noteworthy that D164 belongs to a loop that could potentially interact with the substrate and, if a lysine at residue 234 could commit D164 to a salt bridge, this might result in restricted movement of the loop itself and impact subsequent interactions with substrate.

Although many synonymous variations were demonstrated among different strains within B. cenocepacia subgroups IIIA and IIIB, phylogenetic analysis demonstrated that B. cepacia and B. stabilis each have a unique structural variant of llpE differing from each other and B. cenocepacia subgroups by at least a single non-synonymous variation. It appears that DNA sequence conservation of llpE within a genomovar or subgroup is not maintained, as we observed that the within-group variation is more diverse than the between-group variation. Rather, it is the corresponding protein sequence that is conserved within a genomovar or subgroup, with at least one non-synonymous mutation between any two genomovar groups or subgroups, but none within each of them. Synonymous changes are considered to be accumulating at random in a species, although at a constant rate for a particular gene, thereby acting as a molecular clock. The IIIB node (5 strains) has accumulated a much higher number of synonymous mutations (21 substitutions), than the IIIA node (6 strains) that has only two such changes. Thus, the naïve expectation would be that IIIB is evolutionarily older than IIIA. This may not be true, however. In addition to a low sample size that precludes a definitive resolution of age, the IIIA strains, which were chosen from the B. cepacia experimental panel [[6,[7], might be more closely associated as they are all epidemic strains.

The model structure for LlpE was based on the crystal structure of a novel hyper-thermophilic carboxylesterase from the archaeon A. fulgidus. This structure presents typical features of the α/β hydrolase fold including positioning of the putative catalytic triad residues as well as the GXD/SXG signature motif. This family of proteins includes lipases, esterases, carboxypeptidases and haloperoxidases that show a strong structural similarity, sometimes even in the absence of sequence similarity [[0,[1]. These proteins have the potential to hydrolyze several different types of bonds such as those seen in amides, esters and even carbon–carbon bonds. Among the physiologically important enzymes in the α/β hydrolase superfamily, the human hormone sensitive lipase, an LlpE homolog, has a broad substrate specificity that includes acylglycerols, cholesteryl esters, retinyl esters, steroid esters and p-nitrophenyl esters [[2]. Hydrolytic functions of some of these proteins with pathogenic implications include esterase activity of EstA from Aspergillus niger, an opportunistic fungal pathogen [[3], which also has strong homologs in several pathogenic fungal species. Another member of the α/β hydrolase fold family of enzymes, XylF, catalyzes hydrolytic cleavage of a carbon–carbon bond of meta-cleavage products of aromatic compounds [[4]. Human epoxide hydrolases convert epoxides to the more water-soluble and less toxic diols, thereby protecting us from harmful xenobiotic compounds [[5]. BfaE, a detoxifying enzyme isolated from Bacillus subtilis, hydrolyzes and inactivates Brefeldin A, a potent fungal inhibitor of intracellular vesicle-dependent secretory transport and poliovirus RNA replication [[6], while the haloalkane dehalogenase DhlA from Xanthobacter autotrophicus hydrolyzes 1-haloalkanes to corresponding alcohols [[7,[7]. Similar to DhlA, which has a catalytic triad of D124, D260, H289, LlpE is predicted to possess a catalytic triad composed of D132, D226, H256. Instead of a serine, which usually acts as the catalytic nucleophile, DhlA D124 and, presumably, LlpE D124 act as catalytic nucleophiles. All of the above mentioned members of the α/β hydrolase family of proteins participate in detoxification of exogenous or endogenous substrates. This potential for broad substrate specificity of LlpE may not be unusual for a component of a multidrug efflux operon.

Given the overall high rate of recombination in B. cepacia complex, llpE is remarkably conserved among B. cenocepacia isolates, suggesting the importance of its sequence conservation on some functional activities of the species. Independent of llpE, the other three components of the ceo operon, ceoABopcM, were sufficient to impart chloramphenicol and salicylate efflux to a previously susceptible strain of B. cepacia 249–2 [[1]. LlpE is clearly not essential for antibiotic efflux, because llpE deletion does not adversely influence antibiotic susceptibility and homologues have not been identified associated with other efflux operons. Its conservation among cystic fibrosis isolates suggests a role for LlpE in the unique airway milieu; perhaps its presence imparts an advantage to B. cenocepacia allowing it to thrive in conditions that would otherwise prove toxic. As such it may prove to be a novel target for antimicrobial development.


This work was supported by awards to J.L.B. from the Cystic Fibrosis Foundation (BURNS00P0) and the Royalty Research Fund at the University of Washington. S.C. is supported by NIH Grants GM60731 and AI45820.


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
  26. [26].
  27. [27].
  28. [28].
  29. [29].
  30. [30].
  31. [31].
  32. [32].
  33. [33].
  34. [34].
  35. [35].
  36. [36].
  37. [37].
  38. [38].
  39. [39].
  40. [40].
  41. [41].
  42. [42].
  43. [43].
  44. [44].
  45. [45].
  46. [46].
  47. [47].
View Abstract