posted                         

code:

---

1. Introduction
Borrelia burgdorferi, the causative agent of Lyme disease, is the number one vector-borne disease in the USA [1]. It is a multistage disorder that is difficult to diagnose at any stage of the disease as well as being difficult to treat during the later symptoms. However, during the early stages of the disease, treatment with various antibiotics such as amoxicillin or doxycycline is very efficacious [2]. In vitro, B. burgdorferi cells have high minimum inhibitory concentrations (MICs) for some aminoglycosides (e.g. gentamicin, MIC > 16 μg/mL), quinolones (e.g. nalidixic acid, MIC > 10 μg/mL) and first-generation cephalosporins (e.g. cefadroxil, MIC > 11 μg/mL) but are very susceptible to β-lactams, tetracyclines, fluoroquinolones, everninomycins and macrolides, suggesting that several of these antimicrobials would be effective for the treatment of Lyme disease [3], [4], [5], [6], [7], [8], [9], [10] and [11]. However, therapeutic failures have been reported with most of the appropriate antimicrobials. In fact, poor treatment outcomes have caused the removal of some drugs (e.g. erythromycin) from the list of suitable antibiotics [12]. Whether these failures are due to poor pharmacokinetics in vivo or to the emergence of drug-resistant strains remains unclear. Clearly, both possibilities need to be investigated.

To date, there has been only one report of drug-resistant clinical isolates of B. burgdorferi [13]. This could be due to the difficulty in isolating B. burgdorferi from patients during later stages of the disease, making it hard to determine whether drug-resistant strains are a major cause of poor clinical outcomes. With other bacterial pathogens, the emergence of antibiotic-resistant strains due to mutational events or through the acquisition of foreign DNA has been a major problem in effective treatment. The ease with which resistance genes spread throughout bacterial populations via genetic exchange (e.g. conjugation, transduction or natural transformation) is a major contributing factor. Eggers et al. [14] have shown that B. burgdorferi phage ΦBB-1 can transfer genetic markers between different strains, suggesting that beneficial genetic traits can move within B. burgdorferi populations via transduction. However, owing to the specificity and limited host range of bacteriophages, it is unlikely that B. burgdorferi phage ΦBB-1 fosters the transfer of traits between different genera of bacteria.

We report the isolation of low-passage (LP) and high-passage (HP) strains of B. burgdorferi B31 that exhibit macrolide-lincosamide (ML) or macrolide-lincosamide-streptogramin A (MLSA) resistance patterns. Characterisation of the resistance mechanism suggests that it is due to modification of the ribosomes. More importantly, we demonstrate that the erythromycin-resistant (EmR) phenotype was transferred via conjugation from B. burgdorferi to Bacillus subtilis or Enterococcus faecalis.

2. Materials and methods
2.1. Bacterial strains and reagents
The B. burgdorferi, Escherichia coli, E. faecalis, B. subtilis and Streptococcus pyogenes strains used in this study are listed in Table 1. Borrelia burgdorferi were grown in BSK-II medium [23] at 34 �C under an atmosphere of 3-5% O2, 5% CO2 and 90% N2 and cell numbers were determined by dark field microscopy. Escherichia coli and B. subtilis were grown in Lauria-Bertani (LB) and brain-heart infusion (BHI) medium, respectively. Quinupristin (streptogramin B) and dalfopristin (streptogramin A) were obtained from Aventis Pharma S.A. (Cedex, France). Rabbit serum was purchased from Atlanta Biologicals (Atlanta, GA). [N-methyl-14C]-erythromycin (50 mCi/mmol) was purchased from Amersham Biosciences (Piscataway, NJ). All other reagents were purchased from Sigma Aldrich (St Louis, MO). 

Table 1. 

Bacterial strains used in the study Strain
--------------------------------------------------------------------------------
 Description
--------------------------------------------------------------------------------
 Source
--------------------------------------------------------------------------------
 
Borrelia burgdorferi 
 B31(HP) High-passage, avirulent, passages >100, EmR This study 
 B31(LP) Low-passage, virulent, passages 5, EmR This study 
 B31-5A3 Low-passage, highly virulent Zhang et al. [15] 
 B31-35210 B. burgdorferi type strain ATCC 

 
Escherichia coli 
 BM694 pAT63 ereA, ApR Arthur et al. [16] 
 BM694 pAT72 ereB, ApR Arthur et al. [16] 

 
Enterococcus faecalis 
 JH2-2 FusR, RifR Jacob and Hobbs [17] 
 UV202 FusR, RifR, recA− Yagi and Clewell [18] 
 JH2-2(pAD1) FusR, RifR, Hem-Bac Tomich et al. [19] 
 JH2-2[pAT28 (ermA)] FusR, RifR, SpcR, EmR Trieu-Cuot et al. [20] 
 OG1-SSp StrR, SpcR, TetR M. Roberts 

 
Bacillus subtilis 
 PY79 Prototrophic, rec+, TrR P. Youngman 
 PY1197(pHV1431) Prototrophic, rec+, CmR P. Youngman 
 BD224 recA4 thr-5 trpC2 Dubnau et al. [21] 
 PY79N NalR This study 

 
Streptococcus pyogenes 
 O2C1064 EmR Sutcliffe et al. [22] 


Em, erythromycin; R, resistant; Ap, ampicillin; Fus, fusidic acid; Rif, rifampicin; Hem-Bac, haemolysin, bacteriocin; Spc, spectinomycin; Str, streptomycin; Tet, tetracycline; Tr, trimethoprim; Cm, chloramphenicol; Nal, nalidixic acid.



2.2. Isolation and initial characterisation of EmR B. burgdorferi strains
EmR strains of B. burgdorferi were isolated by plating strains on BSK-II medium containing erythromycin (10, 25 or 50 μg/mL) as described by Samuels et al. [24]. Plates were incubated in a BBL anaerobic GasPak jar without catalyst (Becton Dickinson, Sparks, MD) at 34 �C for 7-14 days. For induction, cultures were grown in BSK-II medium to a cell density of 2 � 107 cells/mL, erythromycin (0.04 μg/mL) was added and the incubation was continued for 24 h. Induced cultures were plated as described above. Resistant colonies were transferred to BSK-II medium containing erythromycin (50 μg/mL), grown to a cell density of 5 � 107 cells/mL and isolates were stored at -80 �C in 50% glycerol.

2.3. Resistance to lincosamides and streptogramins
Erythromycin-sensitive (EmS) and EmR B. burgdorferi isolates were grown in BSK-II medium to a density of ca. 5 � 107 cells/mL and plated on BSK-II medium containing erythromycin (1-100 μg/mL), lincomycin (0.5-15 μg/mL), clindamycin (0.5-10 μg/mL), virginiamycin M (1-15 μg/mL), quinupristin (1-15 μg/mL) or dalfopristin (1-15 μg/mL).

2.4. Erythromycin inactivation assay
Inactivation assays were performed as described by Clancy et al. [25]. BSK-II medium (80 μg/mL erythromycin) was inoculated with isolates B31(LP)-EmR or B31(HP)-EmR and incubated for 60 h at 34 �C. As a positive control, LB medium (80 μg/mL erythromycin) was inoculated with E. coli BM694 harbouring ereA, an erythromycin resistance esterase [16], and incubated at 37 �C for 72 h. Samples were taken at 12-h intervals, cells were removed by centrifugation (10 000 � g, 5 min) and the supernatant was filtered through a 0.22 μm filter (Fisher Scientific, Norcross, GA). Sterile filter disks were placed on plates spread with E. faecalis JH2-2 or E. coli JM109, 5 μL of the culture filtrate was applied to the disks and plates were incubated at 37 �C for 12 h. Zones of inhibition were measured in millimetres. Filtrates from medium with and without erythromycin were used as controls.

2.5. Erythromycin efflux assay
Energy-dependent efflux was assayed as described by Sutcliffe et al. [22] with the following modifications. Cultures of S. pyogenes strain O2C1064 (harbouring a Mef(A) efflux determinant), B. burgdorferi strain B31-EmS and isolate B31-EmR were grown to a cell density of 5 � 107 cells/mL. Cells were collected by filtration, the filters were washed three times with HEPES buffer (50 mM NaCl, 20 mM HEPES, pH 7.4) and cells were counted in a Beckman LS 6500 scintillation counter (Beckman Instruments, Fullerton, CA).

2.6. Binding of [14C]erythromycin to B. burgdorferi ribosomes
Ribosomes were isolated from EmS or EmR B. burgdorferi, E. faecalis and B. subtilis isolates using the method of Goldman et al. [26] and the protein concentration was estimated using the method of Whitaker and Granum [27]. Partially purified ribosomes (7.5 μg) were denatured and mixed with 2.5 nmol (0.25 μCi) [14C]erythromycin (Amersham Biosciences). Ribosomes were collected on 0.2 μm nitrocellulose membranes (Bio-Rad, Hercules, CA), washed 10 times with 10 mM Tris-HCl, 5 mM MgCl2 and 150 mM KCl pH 7.2, and radioactivity was determined using a LS 6000 Beckman scintillation counter (Beckman Instruments). The percent [14C]erythromycin bound by resistant ribosomes was calculated using the following equation:

 


2.7. Polymerase chain reaction (PCR) primers and DNA sequencing
All primers were purchased from Sigma/Genosys Biotechnologies (The Woodlands, TX). The sequences of PCR primers used for the detection of different classes of erythromycin resistance determinants were based upon those described by Sutcliffe et al. [28].

2.8. Genetic exchange from B. burgdorferi to B. subtilis or E. faecalis
Donor cells were prepared as follows: B31(LP)-EmR, B31(HP)-EmR and B31-5A3 cells were grown in BSK-II medium (50 μg/mL erythromycin) to a cell density of ca. 1 � 108 cells/mL. Cells were harvested by centrifugation (3000 � g, 10 min) and re-suspended in BSK-II medium to a cell density of ca. 5 � 107 cells/mL. Recipient cells were prepared as follows: E. faecalis JH2-2 or B. subtilis PY79 were grown to an optical density at 600 nm (OD600) of 0.8. Cultures were diluted 1:1 with fresh medium, harvested by centrifugation (9000 � g, 10 min) and re-suspended in BSK-II medium. Donor and recipient cells were mixed at a ratio of 1:1, 1:2 or 1:3, DNaseI was added and the cells were pelleted by centrifugation (6000 � g, 10 min) to promote cell-to-cell contact. The mixtures were incubated at 34 �C for 18-24 h. The cell pellet was suspended and the incubation continued for 2 h in the presence of 0.04 μg/mL erythromycin (E. faecalis) or 0.06 μg/mL lincomycin (B. subtilis). The E. faecalis/B. burgdorferi mating mixture was plated on BHI medium (25 μg/mL erythromycin), whilst the B. subtilis/B. burgdorferi mating mixture was plated on LB medium (25 μg/mL lincomycin, 1 μg/mL erythromycin). Plates were incubated at 37 �C for 1-3 days. Recipient cells were plated to determine the frequency of spontaneous antibiotic resistance. The frequency of transfer was expressed as the number of transconjugants per recipient cell.

To test for possible genetic transfer via transduction, B31(LP)-EmR or B31(HP)-EmR cells were grown as described. Then, 1 mL of the culture was removed, cells were removed by centrifugation (5000 � g, 15 min) and the supernatant was filtered using a 0.22 μm filter (Fisher Scientific). The filtrate was examined by dark field microscopy to ensure that all spirochetes had been removed. The filtrate was added to E. faecalis or B. subtilis (OD600 = 0.8), incubated at 37 �C overnight and the cells were plated as described above.

3. Results
3.1. Isolation of EmR B. burgdorferi
While testing LP and HP B. burgdorferi B31 strains for susceptibility to various antibiotics, erythromycin-resistant B31(HP)-EmR isolates were identified. These isolates were resistant to high levels of erythromycin (>100 μg/mL), whereas susceptibilities to tetracycline, chloramphenicol, kanamycin and gentamicin were very similar to those reported [2], [4], [7], [8], [9], [10], [11] and [29]. Interestingly, no EmR isolates were obtained for strain B31(LP) in initial experiments. Because the avirulent HP isolate was derived from the virulent LP strain, one would expect the parental strain to display a similar resistance pattern. However, B31(LP)-EmR isolates were obtained when 3 � 107 cells/mL were incubated in BSK-II medium containing 0.04 μg/mL erythromycin prior to plating on BSK-II medium containing erythromycin, suggesting that resistance was inducible in this strain. Subsequent analysis of the LP EmR isolates on higher concentrations of erythromycin indicated that they were also resistant to >100 μg/mL erythromycin. Two additional B31(LP) strains (B31-5A3 and B31-35210) were tested for erythromycin susceptibility; however, no spontaneous or inducible erythromycin resistance was detected in these strains.

3.2. Resistance of B. burgdorferi to the lincosamide and streptogramin antibiotics
To test the resistance patterns of B. burgdorferi to the lincosamide and streptogramin antimicrobials, isolates B31(HP)-EmR and B31(LP)-EmR as well as strains B31-5A3-EmS and B31-35210-EmS were plated onto BSK-II medium containing 15 μg/mL lincomycin, 10 μg/mL clindamycin, 15 μg/mL quinupristin (streptogramin B) or 15 μg/mL dalfopristin (streptogramin A). As expected, LP strains B31-5A3 and B31-35210 were sensitive to all antibiotics tested (Table 2). However, isolate B31(LP)-EmR was resistant to lincomycin and clindamycin but was sensitive to both streptogramins A and B. The B31(HP)-EmR isolate was resistant to dalfopristin (streptogramin A) but sensitive to quinupristin (streptogramin B) (Table 2). These data suggest that these isolates (B31(HP)-EmR and B31(LP)-EmR) do not harbour a typical MLSB resistance determinant. Because the binding site in the 50S ribosomal subunit for streptogramin A does not overlap with those for macrolide, lincosamide and streptogramin B antibiotics, isolate B31(HP)-EmR might harbour an additional determinant conferring resistance to streptogramin A. 

Table 2. 

Macrolide-lincosamide-streptogramin (MLS) resistance patterns of Borrelia burgdorferi strain B31 Antibiotic B. burgdorferi strain B31 from different sources (μg/mL)
--------------------------------------------------------------------------------
 

--------------------------------------------------------------------------------
 (HP)-EmR
--------------------------------------------------------------------------------
 (LP)-EmR
--------------------------------------------------------------------------------
 5A3(LP)
--------------------------------------------------------------------------------
 35210
--------------------------------------------------------------------------------
 
Macrolide 
 Erythromycin 100 100 <1a <1 

 
Lincosamides 
 Lincomycin 15 15 <1 <1 
 Clindamycin 10 10 <1 <1 

 
Streptogramins 
 Virginiamycin Mb 15 1 1 <1 
 Quinupristin (B) 1 1 1 N.D. 
 Dalfopristin (A) 25 1 1 N.D. 


HP, high-passage; LP, low-passage; EmR, erythromycin-resistant; N.D., not determined.
a Cells did not grow at an antibiotic concentration of 1 &#956;g/mL.
b >90% streptogramin A.


3.3. Screening B. burgdorferi EmR isolates for erythromycin resistance determinants
PCR was used to screen B. burgdorferi B31(HP)-EmR, B31(LP)-EmR and B31-5A3-EmS for known erythromycin and streptogramin resistance genes (erm, msr, mef, ere, mph and vga). Although not all genes in these groups were tested, no homologues were found based on the primer sets designed by Sutcliffe et al. [28]. Additionally, analysis of the completed genome sequence did not reveal any potential erythromycin resistance determinants, as expected, since the DNA used for B. burgdorferi genome sequencing was isolated from an EmS B31 strain.

3.4. Inactivation of erythromycin
Whilst rRNA methylase erm genes have been found, high levels of erythromycin resistance are more commonly due to macrolide-inactivating enzymes [30]. Therefore, B. burgdorferi isolates B31(HP)-EmR and B31(LP)-EmR were tested for inactivation of erythromycin. B31(HP)-EmR, B31(LP)-EmR and E. coli BM694 (ereA or ereB) were grown in the presence of erythromycin and culture supernatants were collected at 12-h intervals and tested for activity against EmS E. faecalis. No zones of inhibition were detected from E. coli (ereB) or E. coli (ereA) supernatants after 12 h and 36 h, respectively, indicating inactivation of erythromycin. However, the zones of inhibition remained constant at 3 mm for B31(HP)-EmR and B31(LP)-EmR (Fig. 1), indicating that erythromycin had not been inactivated. No zones were produced with B. burgdorferi cells or E. coli cells alone. These experiments indicated that the erythromycin resistance mechanism in B. burgdorferi was not due to inactivation of the antibiotic. 

 Display Full Size version of this image (25K) 


Fig. 1. Erythromycin inactivation. Filtered culture supernatant from Borrelia burgdorferi isolate B31(HP)-EmR, isolate B31(LP)-EmR, Escherichia coli harbouring ereA or E. coli harbouring ereB was tested for its ability to inactivate erythromycin as described in Section 2.4. Samples were collected at 12-h intervals (0, 12, 24, 36, 48 and 60 h) and zones of inhibition (mm) were measured. Standard error is indicated by error bars. HP, high-passage; LP, low-passage; EmR, erythromycin-resistant. 


3.5. Efflux
Another possible mechanism for the observed erythromycin resistance in B. burgdorferi is active transport of erythromycin via an energy-dependent efflux pump. Uncouplers that poison the proton-motive force effectively block export of erythromycin, causing accumulation within cells. It has previously been demonstrated that the uncoupler carbonyl cyanide p-chlorophenylhydrazone (CCCP) blocks the energy-dependent uptake of [54Mn] in B. burgdorferi [31]. To determine whether resistance to macrolides was mediated by an efflux mechanism, B31(LP)-EmR, B31-5A3-EmS and S. pyogenes 02C1064 cultures were assayed for [14C]erythromycin efflux with and without CCCP. Both EmR and EmS B. burgdorferi cells accumulated [14C]erythromycin rapidly and the intracellular levels stabilised within minutes of exposure to labelled antibiotic (Fig. 2A, B, (&#9830;) solid lines). As expected, intracellular levels of [14C]erythromycin were higher in the EmS strain since sensitive ribosomes bind the antibiotic causing intracellular levels to be higher (Fig. 2B, (&#9830;) solid line). Interestingly, addition of CCCP prior to [14C]erythromycin did not affect the intracellular levels of antibiotic in the EmR isolate B31(LP) (Fig. 2A, (&#9633;) dashed line). In contrast, S. pyogenes 02C1064, which harbours an energy-dependent efflux protein, showed rapid uptake of [14C]erythromycin (Fig. 2C, (&#9830;) solid line) and CCCP dramatically affected the accumulation of [14C]erythromycin (Fig. 2C, (&#9633;) dashed line). These data suggest that erythromycin resistance in strain B31(LP)-EmR was not dependent on an efflux mechanism. 

 Display Full Size version of this image (39K) 


Fig. 2. [14C]Erythromycin uptake assays in the presence or absence of carbonyl cyanide p-chlorophenylhydrazone (CCCP). [14C]Erythromycin was added to cells, aliquots were collected at 10-min intervals (0, 10, 20, 30 and 40 min) and the samples were processed as described in Section 2.5. For some samples, CCCP (100 &#956;m) was added 10 min before the addition of [14C]erythromycin. Incorporation of [14C]erythromycin into (A) Borrelia burgdorferi isolate B31(LP)-EmR, (B) B. burgdorferi strain B31-5A3-EmS, (C) EmR Streptococcus pyogenes (efflux pump) or (D) Enterococcus faecalis strain JH2-2 EmR transconjugant was measured with or without the addition of CCCP. Standard error for each sample set is indicated by error bars. LP, low-passage; EmR, erythromycin-resistant; EmS, erythromycin-susceptible; CPM, counts per min. 


3.6. Binding of [14C]erythromycin to B. burgdorferi ribosomes
The pattern of macrolide and lincosamide resistance in EmR B. burgdorferi isolates and data from the inactivation and [14C]erythromycin uptake experiments suggested that the most likely mechanism of erythromycin resistance in B. burgdorferi was modification of the ribosomes. Since modified ribosomes do not bind erythromycin efficiently, B. burgdorferi ribosomes were analysed for [14C]erythromycin binding [26]. Partially purified ribosomes from B31(LP)-EmR, B31-5A3-EmS, E. faecalis JH2-2 and E. faecalis JH2-2-[pAT28(ermA)] were denatured and mixed with 2.5 nmol (0.25 &#956;Ci) [14C]erythromycin. Binding of [14C]erythromycin to ribosomes purified from B31(LP)-EmR was reduced compared with the EmS strain, indicating that ribosomes from the resistant isolate were modified (Fig. 3A, B). [14C]Erythromycin binding to ribosomes isolated from E. faecalis JH2-2-[pAT28(ermA)] [20] was also reduced in comparison with ribosomes isolated from E. faecalis JH2-2 (Fig. 3E, G). These data suggest that ribosomes from the EmR B31 isolate had been modified. 

 Display Full Size version of this image (23K) 


Fig. 3. Binding of [14C]erythromycin to ribosomes. [14C]Erythromycin was mixed with partially purified ribosomes isolated from (A) Borrelia burgdorferi strain B31-5A3-EmS, (B) B. burgdorferi strain B31(LP)-EmR, (C) Bacillus subtilis strain PY79, (D) B. subtilis strain PY79 EmR transconjugant, (E) Enterococcus faecalis JH2-2, (F) E. faecalis strain JH2-2 EmR transconjugant or (G) E. faecalis JH2-2-[pAT28(ermA)] and assayed as described in Section 2.6. Standard error for each sample set is indicated by error bars. EmS, erythromycin-susceptible; EmR, erythromycin-resistant; LP, low-passage. 


3.7. Transfer of erythromycin resistance from B. burgdorferi to E. faecalis and B. subtilis
Frequently, macrolide resistance determinants identified in Gram-positive bacteria [32] are able to transfer between genera via conjugation. To determine whether the B. burgdorferi EmR phenotype could also be transferred from B. burgdorferi to Gram-positive bacteria, EmR B. burgdorferi isolates were mated with EmS strains of E. faecalis and B. subtilis. The frequency of transfer from B31(HP)-EmR to E. faecalis JH2-2 was 4.5 � 10&#8722;5 (transconjugant/recipient), whilst transfer from B31(LP)-EmR to the same recipient was 2.9 � 10&#8722;5 (Table 3). This was at least three orders of magnitude above the background for spontaneous resistance. EmR E. faecalis transconjugants were resistant to high levels (>100 &#956;g/mL) of erythromycin, whilst spontaneous mutants were sensitive to 25 &#956;g/mL of erythromycin (data not shown). Transfer was not detected to E. faecalis UV202(recA&#8722;) [18] from B31(LP)-EmR or B31(HP)-EmR, indicating that the transfer was dependent on recipient recombination functions (Table 3). Interestingly, E. faecalis transconjugants were able to transfer the erythromycin resistance to a second E. faecalis recipient, strain OG1-SSp, at a frequency (1.0-1.6 � 10&#8722;5) similar to that observed between B. burgdorferi and E. faecalis JH2-2. Finally, transfer was reduced dramatically if the E. faecalis recipient contained a resident plasmid (pAD1) [19] (Table 3). No transfer was observed if B. burgdorferi strain B31-5A3 was used as the donor. 

Table 3. 

Transfer frequency of Borrelia burgdorferi erythromycin resistance Donor
--------------------------------------------------------------------------------
 Recipient
--------------------------------------------------------------------------------
 Transfer frequencya
--------------------------------------------------------------------------------
 Spontaneous frequency
--------------------------------------------------------------------------------
 
B31(HP)-EmR Bacillus subtilis PY79 1.2 � 10&#8722;3 10&#8722;8 
 B. subtilis BD224 1.0 � 10&#8722;4 <10&#8722;9 
 Enterococcus faecalis JH2-2 4.5 � 10&#8722;5 5.4 � 10&#8722;8 
 E. faecalis UV202 6.0 � 10&#8722;9 2.5 � 10&#8722;8 
 E. faecalis JH2-2(pAD1) 7.0 � 10&#8722;7 2.3 � 10&#8722;8 

 
Transconjugant from B31(HP)-EmR � PY79 B. subtilis PY79N <10&#8722;9 <10&#8722;9 
 B. subtilis PY1197 <10&#8722;9 <10&#8722;9 

 
Transconjugant from B31(HP)-EmR � JH2-2 E. faecalis OG1-SSp 1.0 � 10&#8722;5 8.7 � 10&#8722;8 
B31(LP)-EmR B. subtilis PY79 1.1 � 10&#8722;2 10&#8722;8 
 B. subtilis BD224 2.5 � 10&#8722;4 <10&#8722;9 
 E. faecalis JH2-2 2.9 � 10&#8722;5 5.4 � 10&#8722;8 
 E. faecalis UV202 1.7 � 10&#8722;7 2.5 � 10&#8722;8 
 E. faecalis JH2-2(pAD1) 3.3 � 10&#8722;7 2.3 � 10&#8722;8 

 
Transconjugant from B31(LP)-EmR � PY79 B. subtilis PY79N <10&#8722;9 <10&#8722;9 
 B. subtilis PY1197 <10&#8722;9 <10&#8722;9 

 
Transconjugant from B31(LP)-EmR � JH2-2 OG1-SSp 1.6 � 10&#8722;5 8.7 � 10&#8722;8 


HP, high-passage; LP, low-passage; EmR, erythromycin-resistant.
a Per recipient.


The frequency of transfer from B. burgdorferi isolates B31(HP)-EmR and B31(LP)-EmR to B. subtilis was two to three orders of magnitude higher than that observed with E. faecalis and five to six orders above background levels of spontaneous resistance (Table 3). As observed in the previous matings, B. subtilis transconjugants were resistant to >100 &#956;g/mL erythromycin. Transfer was reduced by two orders when EmR B. burgdorferi were mated with B. subtilis strain BD224(recA&#8722;) [21] (Table 3), indicating that, unlike E. faecalis, B. subtilis did not require RecA function for efficient transfer. Again unlike E. faecalis, EmR B. subtilis transconjugants were not able to transfer the resistance phenotype to other B. subtilis strains (Table 3). Although we were able to detect high-frequency conjugal transfer of the EmR phenotype to E. faecalis and B. subtilis, we have not been able to demonstrate transfer of the erythromycin resistance between B. burgdorferi EmR and EmS strains owing initially to a lack of reliable selectable markers in the recipient strains. More recently, identification of a transduction system in B. burgdorferi has complicated the analysis of potential conjugal transfer between Borrelia species.

Further analysis of the transconjugants yielded interesting information. First, EmR B. subtilis and E. faecalis transconjugants did not contain any new extrachromosomal elements, suggesting that B. burgdorferi plasmids were not replicating autonomously in recipient strains. Second, pulsed-field gel electrophoresis analysis of chromosomal DNA from EmR transconjugants showed no large changes to any DNA fragments, indicating that only the transfer of small amounts of DNA was required. Finally, ribosomes purified from EmR B. subtilis and E. faecalis transconjugants were protected from [14C]erythromycin binding (Fig. 3D, F). Taken together, these data indicate the mechanism of erythromycin resistance in the transconjugants was the same as that characterised in the B. burgdorferi EmR isolates.

4. Discussion
Use of erythromycin has had mixed results in the treatment of Lyme disease. Previously, this failure has been attributed to poor penetration of tissues by erythromycin and low attainable serum levels in the human host [29] and [33] since in vitro data on the sensitivity of virulent B. burgdorferi to the macrolide antibiotics [29] and [33] indicate that these strains are sensitive to low levels (<0.06 &#956;g/mL) of erythromycin and lincomycin (<0.25 &#956;g/mL). Terekhova et al. [13] identified clinical isolates of B. burgdorferi that exhibited high-level resistance to erythromycin (100-500 &#956;g/mL) and the resistance was inducible in some strains. However, these strains were not tested for susceptibility to MLS antibiotics nor were the mechanisms of resistance tested experimentally. The EmR strains of B. burgdorferi identified in our laboratory were resistant to high levels of erythromycin (>100 &#956;g/mL). Although the resistance patterns in bacteria are commonly macrolide (M), lincosamide (L), macrolide-streptogramin (MS) or MLSB type [30], Hamilton-Miller and Shah [34] reported similar ML and MLSA resistance patterns in Staphylococcus aureus, Staphylococcus epidermidis and Staphylococcus haemolyticus, but the mechanism(s) has not been identified.

Because we have been unsuccessful at cloning the erythromycin resistance determinant or identifying the gene by PCR, we tried to define the mechanism in the EmR B. burgdorferi strains biochemically. However, testing of the EmR B. burgdorferi for the presence of an antibiotic-inactivating enzyme was negative. Similarly, we found no indication that resistance was associated with an efflux system similar to that previously described in S. aureus (e.g. multiprotein system encoded by msrA) [35], Neisseria gonorrhoeae (mtrRCDE system) [36], Pseudomonas aeruginosa (mexAB-oprK system) [37], E. coli [38] or Streptococcus pneumoniae [22]. Interestingly, the B. burgdorferi genome contains an open reading frame encoding a putative multidrug efflux transporter on linear plasmid 28-4 (lp28-4) with 55% identity to a putative tetracycline resistance protein, TetA(P), from Helicobacter pylori [39]. Since all EmS and EmR B. burgdorferi isolates in this study are sensitive to tetracycline (<0.25-2 &#956;g/mL), it seemed unlikely that this gene encoded a tetracycline efflux protein. However, it was possible that this or other proteins were involved in active efflux of erythromycin in EmR B31 isolates. This possibility was investigated using [14C]erythromycin uptake assays in the presence and absence of CCCP. For B. burgdorferi, addition of CCCP in [14C]erythromycin uptake assays did not affect the incorporation of labelled erythromycin in resistant or sensitive cells, suggesting that the mechanism of erythromycin resistance in these cells was not due to an energy-dependent drug efflux pump.

The most common mechanism for resistance to MLS antibiotics is modification of the ribosome, particularly by methylation of 23S rRNA, with 30 different erm genes described (http://faculty.washington.edu/marilynr/). A less common mechanism of ribosome modification involves amino acid changes in key ribosomal proteins that encompass the antibiotic binding site(s) [40], [41], [42], [43] and [44]. To test whether ribosome modification was responsible for the EmR phenotype in B. burgdorferi EmR strains, we partially purified ribosomes from different isolates and assayed for [14C]erythromycin binding. Equimolar amounts of ribosomes isolated from EmR B. burgdorferi strains bound 85% less [14C]erythromycin than EmS ribosomes, strongly suggesting that the ribosomes had been modified. The gene(s) responsible for this modification and erythromycin resistance in B. burgdorferi have not been identified.

The most interesting finding in the study was the ability to transfer the EmR phenotype via conjugation to two Gram-positive bacteria, E. faecalis and B. subtilis. Many of the MLS genes are associated with conjugative elements and can move into a variety of hosts [30]. Most of the characteristics of the transfer of B. burgdorferi MLSA resistance phenotype resemble those of constins or integrating conjugative elements (ICEs). For example, the B. burgdorferi MLSA resistance transferred by conjugation at high frequency to B. subtilis (10&#8722;2-10&#8722;3) and E. faecalis (10&#8722;5-10&#8722;6). Additionally, the absence of plasmid DNA in the E. faecalis or B. subtilis transconjugants suggested that the resistance determinant was probably integrating into the recipient chromosome [30]. Conversely, integration of B. burgdorferi MLSA resistance, unlike most constins that contain genes encoding recombinases, appeared to be partially or completely dependent on recipient recA function after transfer. Transfer decreased by a factor of 102-104 when a recA&#8722; E. faecalis recipient was used in matings, whilst transfer to recA&#8722; B. subtilis was reduced by 10-102. Although unusual, a 20-50-fold lower frequency of transfer of a Vibrio cholerae conjugative element, SXT, into recA&#8722; recipients has been described [44]. The fact that transfer to recA&#8722; B. subtilis was still occurring at a high frequency (10&#8722;4) suggests that one of the numerous recombinases in B. subtilis, such as RecR, YefB, RecG, etc. (TIGR database) might be facilitating integration in that strain. The exact role of RecA in the recipients is unknown.

Both EmR B. burgdorferi isolates B31(HP) and B31(LP) contain numerous plasmids that could be involved in transfer of the B. burgdorferi MLSA resistance determinant. Two observations hint that they may play a role. First, transfer of the EmR phenotype to E. faecalis decreased dramatically when the recipient strain contained the conjugative plasmid pAD1. Whilst inhibition of transfer of conjugal plasmids due to incompatibility with resident plasmids in recipient strains has been described [45], it is rarely observed with conjugal integrating elements [46] and [47]. Thus, it seems plausible that transfer from B. burgdorferi to E. faecalis may involve donor plasmids that are incompatible with some resident E. faecalis plasmids. Second, a homologue to traB from E. faecalis encoding a pheromone shutdown protein has been identified in B. burgdorferi [39]. TraB is involved in transfer of pheromone-responding plasmids in members of the genus Enterococcus [48]. Unfortunately, no other genes encoding potential transfer factors have been identified on B. burgdorferi plasmids. Despite our limited knowledge of the B. burgdorferi MLSA resistance phenotype and the mechanism of transfer to Gram-positive bacteria, its discovery does suggest that interspecies gene transfer may play a role in the evolution of B. burgdorferi. Additionally, it also provides potential tools for developing new genetic systems to study the pathogenesis of Lyme disease. Studies are ongoing to identify the gene(s) conferring resistance to erythromycin in B. burgdorferi as well as those required for efficient transfer. 

Acknowledgments

We thank Phil Youngman and Dan Zeigler for providing B. subtilis strains, Michael Gilmore for E. faecalis strain UV202, Joyce Sutcliffe, Patrice Courvalin and Amelia Tait-Kamradt for providing E. coli (ereA and ereB) and S. pyogenes 02C1064, and Don Clewell for pAD1.


Funding: This research was supported in part by grant AI33501 from the National Institutes of Health, a Research Training Grant BIR-9413235 from the National Science Foundation, and by the Intramural Research Program of the National Institute of Allergy and Infectious Diseases, National Institutes of Health.


Competing interests: None declared.


Ethical approval: Not required. 

References
[1] Centers for Disease Control and Prevention (CDC). Lyme disease--United States, 2003-2005. MMWR Morb Mortal Wkly Rep 2007; 56:573-6.

[2] B.W. Berger, Current aspects of Lyme disease and other Borrelia burgdorferi infections, Dermatol Clin 15 (1997), pp. 247-255. View Record in Scopus | Cited By in Scopus (13)

[3] W.A. Agger, S.M. Callister and D.A. Jobe, In vitro susceptibilities of Borrelia burgdorferi to five oral cephalosporins and ceftriaxone, Antimicrob Agents Chemother 36 (1992), pp. 1788-1790. View Record in Scopus | Cited By in Scopus (24)

[4] B.W. Berger, Treatment of erythema chronicum migrans of Lyme disease, Ann N Y Acad Sci 539 (1988), pp. 346-351. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (6)

[5] R.J. Dattwyler, J.J. Halperin, H. Pass and B.J. Luft, Ceftriaxone as effective therapy in refractory Lyme disease, J Infect Dis 155 (1987), pp. 1322-1325. View Record in Scopus | Cited By in Scopus (28)

[6] R.J. Dattwyler, J.J. Halperin, D.J. Volkman and B.J. Luft, Treatment of late Lyme borreliosis--randomised comparison of ceftriaxone and penicillin, Lancet 1 (1988), pp. 1191-1194. SummaryPlus | Full Text + Links | PDF (619 K) | View Record in Scopus | Cited By in Scopus (104)

[7] R.C. Johnson, Isolation techniques for spirochetes and their sensitivity to antibiotics in vitro and in vivo, Rev Infect Dis 11 (Suppl. 6) (1989), pp. S1505-S1510.

[8] R.C. Johnson, C. Kodner and M. Russell, In vitro and in vivo susceptibility of the Lyme disease spirochete, Borrelia burgdorferi, to four antimicrobial agents, Antimicrob Agents Chemother 31 (1987), pp. 164-167. View Record in Scopus | Cited By in Scopus (35)

[9] S.E. Johnson, G.C. Klein, G.P. Schmid and J.C. Feeley, Susceptibility of the Lyme disease spirochete to seven antimicrobial agents, Yale J Biol Med 57 (1984), pp. 549-553. View Record in Scopus | Cited By in Scopus (16)

[10] V. Preac-Mursic, B. Wilske and G. Schierz, European Borrelia burgdorferi isolated from humans and ticks: culture conditions and antibiotic susceptibility, Zentralbl Bakteriol Mikrobiol Hyg [A] 263 (1986), pp. 112-118. View Record in Scopus | Cited By in Scopus (104)

[11] B. Wretlind, R.C. Johnson, K. Hansen and V. Preac-Mursic, Antibiotic susceptibility of Borrelia burgdorferi in vitro and in animal models, Scand J Infect Dis Suppl 77 (1991), pp. 143-144. View Record in Scopus | Cited By in Scopus (7)

[12] R.J. Dattwyler, E. Grunwaldt and B.J. Luft, Clarithromycin in treatment of early Lyme disease: a pilot study, Antimicrob Agents Chemother 40 (1996), pp. 468-469. View Record in Scopus | Cited By in Scopus (20)

[13] D. Terekhova, M.L. Sartakova, G.P. Wormser, I. Schwartz and F.C. Cabello, Erythromycin resistance in Borrelia burgdorferi, Antimicrob Agents Chemother 46 (2002), pp. 3637-3640. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (15)

[14] C.H. Eggers, B.J. Kimmel, J.L. Bono, A.F. Elias, P. Rosa and D.S. Samuels, Transduction by &#934;BB-1, a bacteriophage of Borrelia burgdorferi, J Bacteriol 183 (2001), pp. 4771-4778. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (27)

[15] J.R. Zhang, J.M. Hardham, A.G. Barbour and S.J. Norris, Antigenic variation in Lyme disease borreliae by promiscuous recombination of VMP-like sequence cassettes, Cell 89 (1997), pp. 275-285. SummaryPlus | Full Text + Links | PDF (393 K) | View Record in Scopus | Cited By in Scopus (272)

[16] M. Arthur, D. Autissier and P. Courvalin, Analysis of the nucleotide sequence of the ereB gene encoding the erythromycin esterase type II, Nucleic Acids Res 14 (1986), pp. 4987-4999. View Record in Scopus | Cited By in Scopus (32)

[17] A.E. Jacob and S.J. Hobbs, Conjugal transfer of plasmid-borne multiple antibiotic resistance in Streptococcus faecalis var. zymogenes, J Bacteriol 117 (1974), pp. 360-372. View Record in Scopus | Cited By in Scopus (197)

[18] Y. Yagi and D.B. Clewell, Recombination-deficient mutant of Streptococcus faecalis, J Bacteriol 143 (1980), pp. 966-970. View Record in Scopus | Cited By in Scopus (48)

[19] P.K. Tomich, F.Y. An, S.P. Damle and D.B. Clewell, Plasmid-related transmissibility and multiple drug resistance in Streptococcus faecalis subsp. zymogenes strain DS16, Antimicrob Agents Chemother 15 (1979), pp. 828-830. View Record in Scopus | Cited By in Scopus (30)

[20] P. Trieu-Cuot, C. Carlier, C. Poyart-Salmeron and P. Courvalin, A pair of mobilizable shuttle vectors conferring resistance to spectinomycin for molecular cloning in Escherichia coli and in gram-positive bacteria, Nucleic Acids Res 18 (1990), p. 4296. View Record in Scopus | Cited By in Scopus (60)

[21] D. Dubnau, R. Davidoff-Abelson, B. Scher and C. Cirigliano, Fate of transforming deoxyribonucleic acid after uptake by competent Bacillus subtilis: phenotypic characterization of radiation-sensitive recombination-deficient mutants, J Bacteriol 114 (1973), pp. 273-286. View Record in Scopus | Cited By in Scopus (6)

[22] J. Sutcliffe, A. Tait-Kamradt and L. Wondrack, Streptococcus pneumoniae and Streptococcus pyogenes resistant to macrolides but sensitive to clindamycin: a common resistance pattern mediated by an efflux system, Antimicrob Agents Chemother 40 (1996), pp. 1817-1824. View Record in Scopus | Cited By in Scopus (407)

[23] A.G. Barbour, Isolation and cultivation of Lyme disease spirochetes, Yale J Biol Med 57 (1984), pp. 521-525. View Record in Scopus | Cited By in Scopus (456)

[24] D.S. Samuels, K.E. Mach and C.F. Garon, Genetic transformation of the Lyme disease agent Borrelia burgdorferi with coumarin-resistant gyrB, J Bacteriol 176 (1994), pp. 6045-6049. View Record in Scopus | Cited By in Scopus (55)

[25] J. Clancy, F. Dib-Hajj, J.W. Petitpas and W. Yuan, Cloning and characterization of a novel macrolide efflux gene, mreA, from Streptococcus agalactiae, Antimicrob Agents Chemother 41 (1997), pp. 2719-2723. View Record in Scopus | Cited By in Scopus (37)

[26] R.C. Goldman, D. Zakula, R. Flamm, J. Beyer and J. Capobianco, Tight binding of clarithromycin, its 14-(R)-hydroxy metabolite, and erythromycin to Helicobacter pylori ribosomes, Antimicrob Agents Chemother 38 (1994), pp. 1496-1500. View Record in Scopus | Cited By in Scopus (30)

[27] J.R. Whitaker and P.E. Granum, An absolute method for protein determination based on difference in absorbance at 235 and 280 nm, Anal Biochem 109 (1980), pp. 156-159. SummaryPlus | Full Text + Links | PDF (294 K) | View Record in Scopus | Cited By in Scopus (167)

[28] J. Sutcliffe, T. Grebe, A. Tait-Kamradt and L. Wondrack, Detection of erythromycin-resistant determinants by PCR, Antimicrob Agents Chemother 40 (1996), pp. 2562-2566. View Record in Scopus | Cited By in Scopus (388)

[29] V. Preac-Mursic, B. Wilske, G. Schierz, G. S�ss and B. Gross, Comparative antimicrobial activity of the new macrolides against Borrelia burgdorferi, Eur J Clin Microbiol Infect Dis 8 (1989), pp. 651-653. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (25)

[30] M.C. Roberts, J. Sutcliffe, P. Courvalin, L.B. Jensen, J. Rood and H. Seppala, Nomenclature for macrolide and macrolide-lincosamide-streptogramin B resistance determinants, Antimicrob Agents Chemother 43 (1999), pp. 2823-2830. View Record in Scopus | Cited By in Scopus (345)

[31] J.E. Posey and F.C. Gherardini, Lack of a role for iron in the Lyme disease pathogen, Science 288 (2000), pp. 1651-1653. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (99)

[32] D.B. Clewell, Bacterial conjugation (1st ed.), Plenum Press, New York, NY (1993).

[33] A.J. Bryskier, J.P. Butzler, H.C. Neu and P.M. Tulkens, Macrolides: chemistry, pharmacology, and clinical uses, Arnette Blackwell, Boston, MA (1993).

[34] J.M. Hamilton-Miller and S. Shah, Patterns of phenotypic resistance to the macrolide-lincosamide-ketolide-streptogramin group of antibiotics in staphylococci, J Antimicrob Chemother 46 (2000), pp. 941-949. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (34)

[35] L. Wondrack, M. Massa, B.V. Yang and J. Sutcliffe, Clinical strain of Staphylococcus aureus inactivates and causes efflux of macrolides, Antimicrob Agents Chemother 40 (1996), pp. 992-998. View Record in Scopus | Cited By in Scopus (70)

[36] K.E. Hagman, W. Pan, B.G. Spratt, J.T. Balthazar, R.C. Judd and W.M. Shafer, Resistance of Neisseria gonorrhoeae to antimicrobial hydrophobic agents is modulated by the mtrRCDE efflux system, Microbiology 141 (1995), pp. 611-622. View Record in Scopus | Cited By in Scopus (133)

[37] K. Poole, K. Krebes, C. McNally and S. Neshat, Multiple antibiotic resistance in Pseudomonas aeruginosa: evidence for involvement of an efflux operon, J Bacteriol 175 (1993), pp. 7363-7372. View Record in Scopus | Cited By in Scopus (321)

[38] N. Kobayashi, K. Nishino and A. Yamaguchi, Novel macrolide-specific ABC-type efflux transporter in Escherichia coli, J Bacteriol 183 (2001), pp. 5639-5644. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (53)

[39] C.M. Fraser, S. Casjens and W.M. Huang et al., Genomic sequence of a Lyme disease spirochaete, Borrelia burgdorferi, Nature 390 (1997), pp. 580-586. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (891)

[40] K. Nagai, P.C. Appelbaum and T.A. Davies et al., Susceptibility to telithromycin in 1,011 Streptococcus pyogenes isolates from 10 central and Eastern European countries, Antimicrob Agents Chemother 46 (2002), pp. 546-549. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (29)

[41] M. Pihlajamaki, J. Kataja and H. Seppala et al., Ribosomal mutations in Streptococcus pneumoniae clinical isolates, Antimicrob Agents Chemother 46 (2002), pp. 654-658. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (49)

[42] A. Tait-Kamradt, T. Davies and P.C. Appelbaum et al., Two new mechanisms of macrolide resistance in clinical strains of Streptococcus pneumoniae from Eastern Europe and North America, Antimicrob Agents Chemother 44 (2000), pp. 3395-3401. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (171)

[43] A. Tait-Kamradt, T. Davies, M. Cronan, M.R. Jacobs, P.C. Appelbaum and J. Sutcliffe, Mutations in 23S rRNA and ribosomal protein L4 account for resistance in pneumococcal strains selected in vitro by macrolide passage, Antimicrob Agents Chemother 44 (2000), pp. 2118-2125. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (170)

[44] M.K. Waldor, H. Tschape and J.J. Mekalanos, A new type of conjugative transposon encodes resistance to sulfamethoxazole, trimethoprim, and streptomycin in Vibrio cholerae O139, J Bacteriol 178 (1996), pp. 4157-4165. View Record in Scopus | Cited By in Scopus (102)

[45] C. Le Bouguenec, T. Horaud, G. Bieth, R. Colimon and C. Dauguet, Translocation of antibiotic resistance markers of a plasmid-free Streptococcus pyogenes (group A) strain into different streptococcal hemolysin plasmids, Mol Gen Genet 194 (1984), pp. 377-387. View Record in Scopus | Cited By in Scopus (7)

[46] V. Burrus, G. Pavlovic, B. Decaris and G. Guedon, Conjugative transposons: the tip of the iceberg, Mol Microbiol 46 (2002), pp. 601-610. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (60)

[47] G. Whittle, N.B. Shoemaker and A.A. Salyers, The role of Bacteroides conjugative transposons in the dissemination of antibiotic resistance genes, Cell Mol Life Sci 59 (2002), pp. 2044-2054. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (22)

[48] S. Fujimoto, H. Tomita, E. Wakamatsu, K. Tanimoto and Y. Ike, Physical mapping of the conjugative bacteriocin plasmid pPD1 of Enterococcus faecalis and identification of the determinant related to the pheromone response, J Bacteriol 177 (1995), pp. 5574-5581. View Record in Scopus | Cited By in Scopus (35) 

---

--------------------

Posts: 6 | From DUKE | Registered: Jun 2006  |  IP: Logged |