An RND-type efflux system in Borrelia burgdorferi is involved in virulence and resistance to antimicrobial compounds.
ABSTRACT Borrelia burgdorferi is remarkable for its ability to thrive in widely different environments due to its ability to infect various organisms. In comparison to enteric Gram-negative bacteria, these spirochetes have only a few transmembrane proteins some of which are thought to play a role in solute and nutrient uptake and excretion of toxic substances. Here, we have identified an outer membrane protein, BesC, which is part of a putative export system comprising the components BesA, BesB and BesC. We show that BesC, a TolC homolog, forms channels in planar lipid bilayers and is involved in antibiotic resistance. A besC knockout was unable to establish infection in mice, signifying the importance of this outer membrane channel in the mammalian host. The biophysical properties of BesC could be explained by a model based on the channel-tunnel structure. We have also generated a structural model of the efflux apparatus showing the putative spatial orientation of BesC with respect to the AcrAB homologs BesAB. We believe that our findings will be helpful in unraveling the pathogenic mechanisms of borreliae as well as in developing novel therapeutic agents aiming to block the function of this secretion apparatus.
- [Show abstract] [Hide abstract]
ABSTRACT: In bacteria such as Pseudomonas aeruginosa and Escherichia coli, tripartite membrane machineries, or pumps, determine the efflux of small noxious molecules, such as detergents, heavy metals, and antibiotics, and the export of large proteins including toxins. They are therefore influential in bacterial survival, particularly during infections caused by multidrug-resistant pathogens. In these tripartite pumps an inner membrane transporter, typically an ATPase or proton antiporter, binds and translocates export or efflux substrates. In cooperation with a periplasmic adaptor protein it recruits and opens a TolC family cell exit duct, which is anchored in the outer membrane and projects across the periplasmic space between inner and outer membranes. Assembled tripartite pumps thus span the entire bacterial cell envelope.Wereview the atomic structures of each of the three pump components and discuss how these have allowed high-resolution views of tripartite pump assembly, operation, and possible inhibition. Expected final online publication date for the Annual Review of Microbiology Volume 67 is September 08, 2013. Please see http://www.annualreviews.org/catalog/pubdates.aspx for revised estimates.Annual review of microbiology 06/2013; · 12.80 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Periplasmic adaptor proteins are essential components of bacterial tripartite multidrug efflux pumps. Here we report the 2.35Å resolution crystal structure of the BesA adaptor from the spirochete Borrelia burgdorferi solved using selenomethionine derivatized protein. BesA shows the archetypal linear, flexible, multi-domain architecture evident among proteobacteria and retains the lipoyl, β-barrel and membrane-proximal domains that interact with the periplasmic domains of the inner membrane transporter. However, it lacks the α-hairpin domain shown to establish extensive coiled-coil interactions with the periplasmic entrance helices of the outer membrane-anchored TolC exit duct. This has implications for the modelling of assembled tripartite efflux pumps.FEBS letters 07/2013; · 3.54 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: The Borrelia burgdorferi outer membrane (OM) contains numerous surface-exposed lipoproteins but a relatively low density of integral OM proteins (OMPs). Few membrane-spanning OMPs of B. burgdorferi have been definitively identified, and none are well characterized structurally. Here, we provide evidence that the borrelial OMP P66, a known adhesin with pore-forming activity, forms a β-barrel in the B. burgdorferi OM. Multiple computer-based algorithms predict that P66 forms a β-barrel with either 22- or 24-transmembrane domains. According to our predicted P66 topology, a lysine residue (K487) known to be sensitive to trypsin cleavage was located within a surface-exposed loop. When we aligned the mature P66 amino acid sequences from B. burgdorferi and B. garinii, we found that K487 was present only in the B. burgdorferi P66 protein sequence. When intact cells from each strain were treated with trypsin, only B. burgdorferi P66 was trypsin-sensitive, indicating that K487 is surface-exposed as predicted. Consistent with this observation, when we inserted a c-Myc tag adjacent to K487 and utilized surface localization immunofluorescence, we detected the loop containing K487 on the surface of B. burgdorferi. P66 was examined by both Triton X-114 phase partitioning and circular dichroism, confirming that the protein is amphiphilic and contains extensive (48%) β-sheet, respectively. Moreover, P66 also was able to incorporate into liposomes and form channels in large unilamellar vesicles. Finally, blue-native PAGE (BN-PAGE) revealed that under non-denaturing conditions P66 is found in large complexes of ∼400 kDa and ∼600 kDa. Outer surface lipoproteins A (OspA) and OspB both co-immunoprecipitate with P66, demonstrating that P66 associates with OspA and OspB in B. burgdorferi. The combined computer-based structural analyses and supporting physicochemical properties of P66 provide a working model to further examine the porin and integrin-binding activities of this OMP as they relate to B. burgdorferi physiology and Lyme disease pathogenesis.Journal of bacteriology 12/2013; · 3.94 Impact Factor
An RND-Type Efflux System in Borrelia burgdorferi Is
Involved in Virulence and Resistance to Antimicrobial
Ignas Bunikis1,2, Katrin Denker3, Yngve O¨stberg1,2¤, Christian Andersen3, Roland Benz3, Sven
1Department of Molecular Biology, Umea ˚ University, Umea ˚, Sweden, 2Laboratory for Molecular Infection Medicine Sweden (MIMS), Umea ˚ University, Umea ˚, Sweden,
3Lehrstuhl fu ¨r Biotechnologie, Biozentrum der Universita ¨t Wu ¨rzburg, Wu ¨rzburg, Germany
Borrelia burgdorferi is remarkable for its ability to thrive in widely different environments due to its ability to infect various
organisms. In comparison to enteric Gram-negative bacteria, these spirochetes have only a few transmembrane proteins
some of which are thought to play a role in solute and nutrient uptake and excretion of toxic substances. Here, we have
identified an outer membrane protein, BesC, which is part of a putative export system comprising the components BesA,
BesB and BesC. We show that BesC, a TolC homolog, forms channels in planar lipid bilayers and is involved in antibiotic
resistance. A besC knockout was unable to establish infection in mice, signifying the importance of this outer membrane
channel in the mammalian host. The biophysical properties of BesC could be explained by a model based on the channel-
tunnel structure. We have also generated a structural model of the efflux apparatus showing the putative spatial orientation
of BesC with respect to the AcrAB homologs BesAB. We believe that our findings will be helpful in unraveling the
pathogenic mechanisms of borreliae as well as in developing novel therapeutic agents aiming to block the function of this
Citation: Bunikis I, Denker K, O¨stberg Y, Andersen C, Benz R, et al. (2008) An RND-Type Efflux System in Borrelia burgdorferi Is Involved in Virulence and Resistance
to Antimicrobial Compounds. PLoS Pathog 4(2): e1000009. doi:10.1371/journal.ppat.1000009
Editor: Michael R. Wessels, Children’s Hospital Boston, United States of America
Received September 21, 2007; Accepted January 16, 2008; Published February 29, 2008
Copyright: ? 2008 Bunikis et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This study was supported by Swedish Research Council grant 07922, the Swedish Council for Environment, Agricultural Sciences and Spatial Planning
grant 23.0161, Swedish Foundation for Strategic Research (SSF), Infection and Vaccinology and MicMan, The Swedish Foundation for International Cooperation in
Research and Higher Education (STINT), and the JC Kempe foundation, EU project Bovac and the DAAD.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: email@example.com
¤ Current address: CBRN Defence and Security, Swedish Defence Research Agency, Umea ˚, Sweden
Borrelia burgdorferi was described as a causative agent of Lyme
borreliosis in the early eighties [1,2]. The bacterium can be
transmitted to humans by the bite of an infected tick of the genus
Ixodes and cause Lyme borreliosis resulting in a wide variety of
clinical manifestations . The bacterium can survive for an
extended time in humans by evading the immune system, and
leading to chronic infection with arthritis, dermatitis or neurobor-
reliosis. To date, antibiotic treatment with tetracycline or b-lactam
is successful in most cases , especially when infection is
diagnosed at an early stage. However, B. burgdorferi has natural
resistance towards several antibiotics such as phosphomycin and
sulfamethoxazole. The mechanism(s) of resistance are unknown
but may include those already described for other bacteria, such as
enzymes, mutation of the antibiotic target, or efflux pumps.
In comparison to enteric Gram-negative bacteria, the density of
spirochetes B. burgdorferi has the highest concentration of transmem-
brane proteins [5,6]. Porins are one class of transmembrane proteins
responsible for solute transport across the outer membrane of Gram-
negative bacteria. Integral outer membrane proteins with functions
similar to porins have been reported for Borrelia [7,8,9,10].
During the course of evolution, bacteria have been exposed to a
variety of toxic compounds such as toxins, endogenous metabolic
end products, and antibiotics. To protect themselves, microor-
ganisms have evolved devices to detoxify and secrete these
substances. Bacterial resistance to many classes of antibiotics is
provided mainly by membrane transporter proteins called drug
efflux pumps , which are part of the multi-drug resistance
(MDR) efflux systems. Five main families of bacterial MDR
transporters have been identified: MF (major facilitator), MATE
(multi-drug and toxic efflux), ABC (ATP binding cassette), SMR
(small multi-drug resistance), and RND (resistance-nodulation-
division) . RND transporters exist in all kingdoms of living
organisms, but seem to be involved in drug resistance especially in
Gram-negative bacteria where they export toxic substances across
both membranes of the cell envelope in a single energy-coupled
step . These efflux pumps are made of three components: a
cytoplasmic membrane export system that acts as an energy-
dependent pump, a membrane fusion protein (MFP), and an outer
membrane factor (OMF) [13,14]. The major antibiotic efflux
activity of this type in Escherichia coli is mediated by the tripartite
multi-drug resistance pump AcrAB-TolC [15,16] which transports
substrates from the cell into the external medium, bypassing the
periplasm and the outer membrane . This complex consists of
PLoS Pathogens | www.plospathogens.org12008 | Volume 4 | Issue 2 | e1000009
the inner membrane translocase AcrB which is thought to be a
proton transporter [18,19] belonging to the resistance-nodulation-
division (RND) family of proteins , the outer membrane
channel TolC, which is an OMF , and a periplasmic linker
protein, AcrA, which is a member of the membrane fusion protein
(MFP) superfamily . TolC forms trimers, in which each
monomer contributes four b-strands to a single channel-forming
unit. TolC homologues are ubiquitous among Gram-negative
bacteria, and thus far nearly a hundred have been identified .
In this study, we identified a channel-forming activity
corresponding to a TolC homolog, BesC, in the B. burgdorferi
outer membrane. We show that the BesC protein is necessary for
B. burgdorferi to establish infection in mice and is involved in
antibiotic resistance. Furthermore, we determined the biophysical
properties of the channel formed by BesC and generated a model
of the putative efflux apparatus.
In silico identification of a Borrelia TolC homolog
In a comprehensive database search for outer membrane porins
of different Gram-negative bacteria Yen et al. predicted that the
Borrelia burgdorferi B31 genome harbors a putative member of an
outer membrane factor (OMF) family hypothetical protein
designated BB0142 . We analyzed the amino acid sequence
of this protein using an NCBI conserved domain search which
revealed significant similarity (E value = 5222) to proteins of the
outer membrane efflux protein (OEP) family, including the E. coli
outer membrane protein TolC . This is an important
indication that BB0142 is a TolC homologue. The Borrelia genome
contains only one gene encoding a protein belonging to the OEP
family. Investigation of the genes directly flanking bb0142 showed
that bb0140 codes for AcrB-like protein, an inner membrane
transporter of the RND (resistance, nodulation, cell-division)
family, and bb0141 codes for an AcrA-like protein, an adaptor
protein also known as membrane fusion protein. Homologs of
these proteins form tripartite multi-drug efflux pumps in many
Gram-negative bacteria. Therefore, we propose renaming these
genes besA (bb0141), besB (bb0140) and besC (bb0142) for Borrelia
efflux system proteins A, B and C.
besB, besA, and besC are co-transcribed
According to the B. burgdorferi genome sequence, the genes besB
(bb0140), besA (bb0141) and besC (bb0142) are oriented in the same
direction, and separated by 18 and 8 bp, respectively. In order to
determine whether the three genes are transcribed as a single
transcript, we analyzed RNA from low-passage strain B31 by RT-
PCR using primers specific for different genes (Table S1). The
primer pair spanning junction of besB and besA amplified a
fragment (lane 4 in Figure 1) demonstrating that both genes are
located on the same transcript. Similar results were obtained for
primer pair spanning junction of besA and besC genes (lane 5 in
Figure 1). Therefore we conclude that all three genes are
transcribed together. This was further confirmed by amplifying a
product spanning besB, besA and besC using primers specific for besB
and besC (lane 6 in Figure 1). Negative control reactions, in which
reverse transcriptase was omitted, yielded no such product (lane 3
in Figure 1), indicating that the products were derived from RNA
rather than contaminating DNA. Additionally, besC-specific
primers were used as a positive control (lane 7 in Figure 1). These
data show that besB, besA and besC are transcribed as a single
Inactivation and complementation of the besC gene in B.
In order to investigate the potential involvement of BesC in
antibiotic resistance and virulence of B. burgdorferi, we constructed
a besC mutant. A streptomycin resistance gene was inserted into
the besC gene (Figure 2A) of low-passage infectious B. burgdorferi
strain 5A4NP1 by electroporation with plasmid pOK-besC::str as
described in Materials and Methods. Plasmid pCOMP (Figure 2A)
was used to complement besC mutant. Strains were analyzed by
PCR to confirm inactivation and complementation of besC
The expression of BesC protein in strain 5A4NP1, the besC
mutant and the complemented strain was analyzed by immuno-
blotting with antiserum raised against the recombinant construct
of BesC. As expected, BesC was present in the wild-type and
complemented strains, but not in the mutant besC::str (Figure 2C).
Due to the finding that besC is co-transcribed with besA and besB,
membranes were further probed with antibodies raised against
recombinant parts of BesA and BesB to determine whether protein
expression is also linked. The putative membrane fusion protein
Figure 1. Schematic representation of the putative besB-besA-
besC operon in Borrelia burgdorferi genome and results from the
reverse transcription PCR. Lane 1 and 8; molecular weight markers,
lane 2; PCR reaction without template using primer pair c2 and c3
served as negative control, lane 3; primer pair c2 and c3 in ordinary PCR
reaction using RNA as template served as negative control for DNA
contamination, lane 4; primer pair b1 and a1, lane 5; primer pair a2 and
c1, lane 6; primer pair b1 and c1, lane 7; primer pair c2 and c3 served as
Lyme disease is caused by infection with the spirochete
Borrelia burgdorferi. These spirochetes cycle between
Ixodes ticks and vertebrate reservoirs, mainly rodents, but
also birds. Previous studies have revealed major differenc-
es in the B. burgdorferi cell envelope structure and
membrane composition compared to those of other
bacteria. Proteins embedded in the bacterial membranes
fulfill a number of tasks that are crucial for bacterial cells,
such as solute and protein transport, as well as signal
transduction, and interaction with other cells. Microorgan-
isms have evolved mechanisms to protect themselves
against harmful substances and secrete these through
efflux pumps. So far, little is known about mechanisms of
drug efflux systems in borreliae. Herein we identified an
outer membrane channel forming protein important for B.
burgdorferi to cause infection in mice and that also is
involved in antibiotic resistance. We believe that this work
will be helpful to understand the mechanisms underlying
borreliae infection biology as well as in developing new
therapeutic agents aiming to block this multi drug efflux
Type 1 Secretion System in Borrelia
PLoS Pathogens | www.plospathogens.org22008 | Volume 4 | Issue 2 | e1000009
BesA and putative inner membrane transporter BesB were
expressed in low-passage infectious strain 5A4NP1, as well as in
the complemented strains, but neither protein could be detected in
the besC::str mutant (Figure 2C). This observation further
confirmed the RT-PCR results indicating that the three genes
are transcriptionally linked.
No phenotypic differences between 5A4NP1 and besC::str
mutant were observed when the morphology of the spirochetes
was compared by microscopy. Growth curves were determined for
the 5A4NP1, besC::str and the complemented strains under normal
growth conditions in liquid medium by counting the spirochetes in
a Petroff-Hausser chamber daily. No significant effect on growth
could be observed (data not shown).
Involvement of BesC in antibiotic resistance
It has been shown previously that the outer membrane
components are essential for functionality of multi-drug efflux
pumps . Therefore susceptibilities of B. burgdorferi strain
5A4NP1, besC::str mutant strain and the complemented strain to
different antimicrobials were tested in vitro. In summary, both the
MIC (minimal inhibitory concentration) and MBC (minimal
borreliacidal concentration) values of the different compound
tested for B. burgdorferi carrying inactive besC were 2- to 64-fold
lower as compared to the parental strain 5A4NP1 (Table 1). The
MIC and MBC values were essentially the same for the parental
strain and complemented mutant (Table 1). Our results therefore
suggest that BesC is involved in antibiotic resistance in B.
BesC is essential for mouse infection
To investigate the influence of besC inactivation on B. burgdorferi
virulence, animal infectivity experiments were performed. C3H/
HeN mice were infected with strain 5A4NP1, besC::str and
complemented strain besC::str + besC. After two weeks heart,
bladder, knee and ear were transferred to fresh BSKII media.
Cultures were monitored for 4 weeks for spirochete growth. The
findings are summarized in Table 2. As expected, cultures from
positive control mice contained vigorously growing bacteria. In
cultures from mice infected with besC mutant no bacterial growth
occurred, suggesting that BesC is important for Borrelia to establish
infection in mice. Furthermore, the complemented strain could
infect mice and was re-isolated from all cultured organs. In vitro
propagation of B. burgdorferi strains can result in loss of plasmids;
therefore both pre- and post-infection strains were examined for
plasmid content in order to confirm that loss of infectivity of
besC::str strain was not due to loss of important plasmids but due to
inactivation of besC. All strains contained an intact set of
extragenomic elements (data not shown). The presence of the
shuttle vector used for complementation in strain besC::str + besC
recovered from infected mice organs was confirmed by PCR as
Figure 2. Characterization of the besC mutant strains. (A) Schematic representation of besC, becA and besB genes in the Borrelia burgdorferi B31
chromosome, insertion of the aadA gene cassette by homologous recombination and complementation plasmid. Arrows indicate the relative
positions of the oligonucleotides used. The diagram is not drawn to scale. (B) PCR analysis of the wild-type strain, the resulting besC mutant and the
complemented strain using primer pairs specific either only for the besC and aadA genes; or one primer specific for besA and the other for aadA gene.
The PCR was also used to confirm the presence of the shuttle vector in the complemented strain using one primer specific for besC and another
specific for the shuttle vector. (C) Immunoblotting using antiserum raised against BesA, BesB and BesC to determine the presence or absence of these
proteins in wild-type, besC mutant and the complemented strain.
Type 1 Secretion System in Borrelia
PLoS Pathogens | www.plospathogens.org32008 | Volume 4 | Issue 2 | e1000009
well as by transforming total DNA from this strain into E. coli cells
and isolating gentamycin-resistant colonies. Subsequent screening
of colonies by PCR using primers specific for the shuttle vector
confirmed the presence of the pCOMP plasmid (data not shown).
Identification of BesC channel-forming activity in the B.
burgdorferi outer membrane fraction
P66 is a channel-forming protein with a single channel
conductance of 11 nS in 1 M KCl, making it the largest outer
membrane channel present in B. burgdorferi . Absence of the
giant P66 outer membrane channel represents a considerable
advantage for purification of other B. burgdorferi channel-forming
proteins . Therefore, strain p66::str was used to separate outer
membrane proteins by anion exchange chromatography. Proteins
were eluted with buffer supplemented with increasing concentra-
tions of NaCl and analyzed in lipid bilayer assays to detect
channel-forming activity. The chromatogram taken at 280 nm
showed a small protein peak for the fractions eluted at 250 mM
NaCl (data not shown), which showed uniform channel-forming
activity with a single-channel conductance of about 300 pS. To
analyze its protein content, the active fraction was precipitated and
separated by SDS-PAGE. Silver stained protein bands were
analyzed by mass spectrometry and identified by peptide mass
fingerprinting (Figure S1 and Table S2). A protein band with an
apparent molecular mass of 48 kDa was identified as BesC, which
was verified by Western blot (data not shown). Addition of
polyclonal BesC antibodies to the FPLC fraction inhibited
channel-forming activity in a lipid bilayer assay, further confirm-
ing that BesC is responsible for the channel-forming activity.
To further confirm our findings we analyzed outer membrane
fraction of p66::str/besC::kan mutant strain separated by anion
exchange chromatography as described above. No channel-
forming activity with a single-channel conductance of 300 pS
could be observed.
Channel-forming properties of purified BesC
To analyze the channels formed by BesC, single-channel
experiments were performed with purified protein. FPLC fractions
containing BesC resulted in a step-wise increase of the membrane
conductance in a lipid bilayer assay. Figure 3A shows a single-
channel recording of a lipid bilayer membrane in the presence of a
very low concentration of BesC at a membrane voltage of 20 mV.
Figure 3B shows a histogram of 122 conductance steps observed in
single-channel experiments with BesC-containing FPLC fractions.
This protein forms obvious channels with an average single-
channel conductance of 300 pS in 1 M KCl. Interestingly the
300 pS pore did not display any voltage dependence even at
voltages as high as 6 150 mV (data not shown).
The channels formed by BesC were permeable to a variety of
different ions. The conductance of the 300 pS channel was found
to be dependent both on the type of electrolyte and its
concentration (Table 3). The KCl concentration was varied from
0.1 to 3 M and the conductance behaved as a linear function of
the electrolyte concentration, meaning that the channel does not
contain point charges in or near the channel mouth. Replacement
of chloride against the less mobile acetate had a strong effect on
the single-channel conductance, which decreased by 50% from
300 pS for 1 M KCl to 150 pS for 1 M potassium acetate.
Exchange of the cation from K+to Li+had, in contrast to this, a
relatively small effect: the single-channel conductance decreased
only from 300 pS (1 M KCl) to 250 pS (1 M LiCl). This result
demonstrates that the 300 pS channel shows some preference for
anions over cations.
Selectivity of the BesC channel
Zero-current membrane potential experiments were performed
to analyze the ion selectivity. Table 4 shows the results of
measurements taken in the presence of 5-fold salt gradients of
KCl, LiCl and potassium acetate. After insertion of 100 to 1000
channels into the phosphatidylcholine membrane, the salt
concentration on one side of the membrane was raised from 100
to 500 mM by addition of 3 M salt solution. The aqueous phase
was stirred for equilibration and 10 min after increasing the salt
gradient the zero-current potential across the membrane was
measured. For potassium acetate the potential was found to be
positive on the more diluted side of the membrane, whereas it was
found to be negative for LiCl on the same side. The zero-current
membrane potential was close to zero for KCl, which means that
the ion permeability through the BesC pore follows the aqueous
mobility of the ions. Analysis of the zero-current membrane
potential using the Goldman Hodgkin Katz equation  revealed
a permeability Pcation/Panion of 0.6 (LiCl), 0.9 (KCl) and 2.4
(potassium acetate), respectively.
Table 1. In vitro antibiotic susceptibility of B. burgdorferi.
MICMBCMIC MBC MICMBCMICMBCMIC MBCMIC MBCMIC MBC
WT (5A4NP1) 0.6320 0.16 1.250.312.50 0.0160.1250.050.781.56 6.25 781.25781.25
besC::str0.3150.080.310.040.16 0.0020.031 0.010.200.050.3912.2112.21
besC::str + besC0.6320 0.161.250.312.500.0160.1250.050.781.566.25781.25781.25
aStrains used are described in Materials and Methods.
bThe MIC and MBC values for the antimicrobial agents are in mg.
Table 2. Mouse infection study using B. burgdorferi wild-type,
besC::str mutant and complemented strains.
StrainRe-isolation from tissues
No. of mice infected/
besC::str + besC7/7 7/77/77/77/7
Type 1 Secretion System in Borrelia
PLoS Pathogens | www.plospathogens.org42008 | Volume 4 | Issue 2 | e1000009
A structural model explains the electrophysiological
properties of BesC
To further analyze the electrophysiological properties of BesC in
comparison with the best characterized channel tunnel TolC of E.
coli, we have modeled the structure of BesC. A sequence alignment
with sequences of structurally known channel tunnels [13,28,29]
revealed that BesC shares important structural determinants of the
channel tunnel family (http://npsa-pbil.ibcp.fr/) (Text S1). For
example, BesC contains prolines at positions P44 and P251 that are
highly conserved within the TolC protein family. These proline
residues are strictly required to accommodate the abrupt turn that
links the b-barrel to the a-helical motifs. Similarly, the glycine
residuesatpositions151 and 370 arealsoconservedand aresituated
in the turns near the closed end of the periplasmic domain.
However, the main differences are found at the amino- and
carboxyterminal ends of the proteins. The carboxy-terminus of
BesC is the shortest. It ends directly after helix 8 and does not form
an extra structure within the equatorial domain as the 19 or 66
residues of the carboxy-termini of OprM or TolC, respectively. The
length of the amino-terminus outside the tunnel structure is for
BesC in the same range as that of TolC (13 and 10 residues,
respectively. In contrast to these the OprM amino-terminal end is
61 residues long and possess an acylation site, which anchors it in
the outer membrane. There are also minor variations in the length
of the extracellular loops. It is noteworthy however, that these
variations as well as those of the equatorial domain also have been
reported for other TolC homologues .
To explain our biophysical data, a special focus was put on
residues lining the periplasmic entrance, which are known to have a
major influence on the electrophysiological properties of channel-
tunnels . In comparison to TolC of E. coli, which forms pores of
80 pS in 1 M KCl with a high preference for cations due to six
aspartate residues (Asp371 and Asp374) lining the tunnel entrance
(Figure 4A), the pores formed by BesC are almost non-selective.
Looking at the modeled BesC tunnel entrance, it becomes apparent
that the opening is slightly wider than that of TolC, explaining the
higher single channel conductance. Furthermore, the charges lining
the channel entrance are balanced. There are two oppositely
charged residues per monomer, Asp363 and Lys366, which could
explain the low ion selectivity of BesC (Figure 4B).
The adaptor protein of the Borrelia multi-drug efflux
pump is atypical
Sequence alignments with homologs of the other two proteins,
which are part of the Borrelia drug efflux pump, showed no major
Figure 3. Channel-forming activities of FPLC purified BesC. (A)
Single-channel recording of a diphytanoyl phosphatidylcholine/n-
decane membrane observed in the presence of 100 ng/ml of fraction
35 of the MonoQ-FPLC of the B-fraction of B. burgdorferi B31-A p66::str.
The aqueous phase contained 1 M KCl; Vm = 20 mV; T = 20uC. (B)
Histogram of the probability P(G) for the occurrence of a given
conductivity unit observed with membranes formed of 1% diphytanoyl
phosphatidylcholine/n-decane in the presence of fraction 35 of the
MonoQ-FPLC of the B-fraction of B. burgdorferi B31-A p66::str. P(G) is the
probability that a given conductance increment G is observed in the
single-channel experiments. It was calculated by dividing the number of
fluctuations with a given conductance increment by the total number
of conductance fluctuations. The average single-channel conductance
for 122 single-channel events was 300 pS.
Table 3. Average single channel conductance, G, of the BesC
channels in different salt solutions.
Average single channel
KCH3COO, pH 71150
The membranes were formed from 1% diphytanoylphoshatidylcholine
dissolved in n-decane. The pH of the aqueous salt solutions was 6 if not
indicated otherwise. The single-channel conductance is given as the mean of at
least 100 single events. The applied voltage was 20 mV and the temperature
Table 4. Zero-current membrane potential, Vmof
diphytanoylphoshatidylcholine/n-decane membranes in the
presence of BesC measured for a 10-fold gradient of different
SaltVma, mVPermeability ratio Pcation/Panion
KCH3COO, pH 7102.4
aVmis defined as the difference between the potential at the diluted side
(100 mM) and the potential at the concentrated side (500 mM). The pH of the
aqueous salt solution was 6 unless otherwise indicated. t = 20uC. Pcation/Panion
was calculated with the Goldman-Hodgkin-Katz equation  from at least
three individual experiments.
Type 1 Secretion System in Borrelia
PLoS Pathogens | www.plospathogens.org5 2008 | Volume 4 | Issue 2 | e1000009
differences for the RND transporter BesB compared to other
members of the family (Text S1). There is however a striking
difference in BesA compared to other adaptor proteins involved in
multi-drug efflux. The sequence alignment reveals that the alpha-
helical domain, which is located between the two halves of the
highly conserved lipoyl domain, is missing in BesA. The loop,
which connects the two halves is just nine residues long instead of
the 61 or 75 residues found in the proteins MexA of P. aeruginosa or
AcrA of E. coli, respectively. This means that the long coiled coil
domain, which is suggested to stabilize the contact to the channel-
tunnel in the assembled complex of AcrABTolC or MexABOprM,
is missing . Thus, in contrast to other efflux pumps
investigated so far with extensive contact sites between the inner
and outer membrane components, this contact is restricted to
head-to-tail interactions between the periplasmic end of BesC at
one site and the uppermost part of the inner membrane complex
formed by BesA and BesB in the Borrelia efflux pump (Figure 4C).
In this study, we characterized a putative RND-type efflux
system in Borrelia burgdorferi. Based on the sequence homology to
other multi-drug efflux systems, we conclude that the besB, besA,
and besC genes encode an efflux system in B. burgdorferi. This was
further supported by analysis of BesC channel-forming activities in
a black lipid bilayer assay.
b-lactam antibiotics, macrolides, and tetracyclines are generally
recommended for stage-dependent therapy of Lyme borreliosis,
and borreliae are resistant to aminoglycosides and quinolones,
such as ciprofloxacin acid and ofloxacin . Existing evidence
indicates that the possible heterogeneity of B. burgdorferi may enable
certain isolates to evade antimicrobial therapy and may account
for the subsequent relapses suffered by some patients [33,34,35].
The mechanisms and actual proteins involved in Borrelia resistance
to different antimicrobial agents have not been elucidated.
Therefore our finding of increased Borrelia susceptibility to several
antibiotics (Table 1) due to inactivation of BesC, the homolog of E.
coli TolC, is the first proof that Borrelia possesses an active efflux
system which might be responsible for multi-drug resistance. This
idea can be supported by studies of E. coli TolC and its role in
efflux of different substances including penicillins and tetracyclines
The emergence of active efflux as a major causative factor in
antibiotic resistance has been one of the most significant trends in
anti-infective chemotherapy over the last decade and strategies to
identify efflux pump inhibitors are in progress . For these
reasons, the identification and the characterization of such an
efflux system in B. burgdorferi is very important for understanding
the pathogenicity of this organism.
There is accumulating evidence that efflux pumps conferring
clinically relevant antibiotic resistance are important for bacterial
pathogenicity. This was further strengthened in a recent study by
Gil and coworkers that the deletion of a TolC ortholog in
Francisella tularensis affected its role both in virulence and in
antibiotic resistance . It is possible that in some species efflux
pumps exporting antimicrobial agents also are important for
colonization and infection of human and animal cells [38,39].
Several studies have shown that lack of efflux-pump expression by
a Gram-negative bacterium has a deleterious effect on the ability
of the bacterium to be pathogenic in animal models [40,41]. Some
bacterial efflux pumps export not only antibiotics and other
substances, but also host-derived antimicrobial agents . This
finding has led to the suggestion that the physiological role of these
systems is evasion of such naturally produced molecules, thereby
allowing the bacterium to survive in its ecological niche .
In Gram-negative bacteria, all three components of the efflux
systems are often encoded within the same gene cluster, as is the
Figure 4. Comparison of the periplasmic tunnel entrance of E. coli TolC with the modeled structure of Borrelia BesC and a model of
the Borrelia efflux pump. The backbone of the peptide chains of the three monomers are colored differently. Side chains are omitted except those
lining the tunnel entrance. These are D371 and D374 in TolC (A) and D363 and K366 in BesC (B) or those involved in forming the circular network in
TolC - D153, Y362, and R367. (C) The channel-tunnel BesC is colored red, the adaptor protein BesA is shown in yellow and the RND transporter in
green. Structures of all three components are modeled and formed according to existing models [28,82,83]. The model of the adaptor protein BesA
shows just residues 38–219 of the mature chain. The remaining parts of the protein are not modeled because of the missing template. In this model
the complex is comprised of three adaptor protein protomers. It should be mentioned that there are other models proposed suggesting six adaptor
protein protomers per efflux apparatus [31,59]. Note that the adaptor proteins do not have the alpha-helical domain, which is thought to interact
with the helices of the tunnel region of the outer membrane component in other bacterial efflux pumps.
Type 1 Secretion System in Borrelia
PLoS Pathogens | www.plospathogens.org62008 | Volume 4 | Issue 2 | e1000009
case for mexAB-oprM from Pseudomonas aeruginosa . However, in
some efflux systems, the gene encoding the OMF component is
not in the same gene cluster as the other two components. In B.
burgdorferi, the genes coding for the three components are clustered,
and form one transcriptional unit. Furthermore, in several Gram-
negative bacteria, more than one TolC homologue and multiple
multi-drug efflux systems are present . This is not the case in
B. burgdorferi. Here, the besABC system is the only multi-drug efflux
pump identified thus far in the genome. In addition to the TolC
efflux protein, multi-drug efflux and type I secretion systems
require a periplasmic adaptor protein and an energy-providing
Functional multi-drug efflux pumps need several components.
BesC needs to interact with the putative inner membrane complex
formed by the AcrB homologue BesB, which serves as an inner
membrane transporter, and the AcrA homologue BesA, a
periplasmic adaptor protein. The genes coding for these three
proteins are likely to form an operon within the chromosome of B.
burgdorferi. These three components are necessary to form a
transport system spanning the two membranes for pumping out
noxious compounds from the cytoplasm [25,46]. The special
characteristic of the Borrelia adaptor protein is the absence of the
hairpin domain. Several experiments with E. coli and P. aeruginosa
efflux pumps support the idea that the hairpin domain is involved
in stabilization of the contact between inner and outer membrane
components [31,47]. In the case of the Borrelia efflux pump, the
modeled structures of the three Borrelia components suggest that
the contact zone is restricted to the head-to-tail contact between
the BesC channel-tunnel and the BesB/BesA complex. In this
interaction charged residues might play a role. There are
negatively charged residues on the tip of the tunnel structure
and positively charged residues at loops forming the rim of the
funnel structure of BesB. Also the nine-residue-long loop replacing
the helical domain in BesA contains two positively charged
residues, which might be involved in the interaction.
It is widely believed that the opening of the tunnel entrance is
triggered by the interaction with the inner membrane complex. In
the efflux pumps investigated and modeled by now it is obvious
that the coiled coil domains of the adaptor proteins provide a large
interface for such interaction. However, biochemical data of the E.
coli efflux pump also provide evidence that loops of the upper rim
of the inner membrane transporter AcrB are in contact with the
channel tunnel and might also trigger and stabilize the tunnel
opening [31,47,48]. It must be assumed that in the case of the
Borrelia efflux pump this interaction is sufficient to trigger and
stabilize opening of BesC. One might speculate that compared to
the more tightly constricted periplasmic opening of OprM, which
opening requires larger conformational changes and probably
more energy input, possibly provided by the stronger interaction
due to the coiled coil domains of the adaptor proteins, the already
wider tunnel entrance of BesC requires less conformational
changes and less interaction for the transition into the open state.
One can assume that the lack of the coiled coil domain of the
BesA protein concomitant with a much smaller interaction site
between the inner membrane complex and the outer membrane
component results in a less stable assembly compared to drug
efflux pumps like AcrABTolC or MexABOprM. From an
evolutionary point of view the Borrelia efflux pump presents an
archetypical form of this apparatus because the adaptor protein
lacking the coiled coil domain resembles most the potential
progenitor, the lipoyl-domain of acetyl transferase proteins .
This is consistent with the fact that Spirochaete present a very
early branch in evolution of bacteria .
Separation of the outer membrane proteins from a p66 knock
out strain by anionic exchange chromatography resulted in a
fraction, which showed a uniform channel-forming activity in the
black lipid bilayer assay with a single-channel conductance of
300 pS in 1 M KCl. By mass spectrometry using peptide mass
fingerprints  we identified the protein BesC as the sole protein
with a putative pore-forming function. A BLAST search and a
NCBI conserved domain search identified BesC as a homologue of
E. coli TolC. TolC is the outer membrane component of type I
secretion systems and multi-drug efflux pumps . By itself, TolC
forms a channel with a single-channel conductance of 80 pS in
1 M KCl, which is almost four-fold smaller as that for the channel
in the isolated FPLC fraction [13,45,52,53]. TolC is a homotrimer
with a 140 A˚long cannon-shaped structure and each of the three
monomers contributes four b-strands to form a single 40 A˚long b-
barrel, the so-called channel domain, which is anchored in the
outer membrane . The b-barrel is extended by a 100 A˚long
tunnel domain formed exclusively by a-helices. The alignment of
BesC with TolC of E. coli and OprM of P. aeruginosa (Text S1)
reveals conserved residues, which are important for the correct
folding of the channel-tunnel family. Among these are proline and
glycine residues, which are important for the transitions between
the b-sheets of the channel and the a-helices of the tunnel . An
immunoblot analysis with BesC antibodies showed that the protein
is present in the FPLC fraction, which also is active in the black
lipid bilayer (data not shown).
For TolC of E. coli it is known that residues lining the
periplasmic entrance are critical for the electrophysiological
behavior. The aspartate ring consisting of six aspartates residues
at the TolC tunnel entrance from the periplasm is responsible for
its high cation selectivity . In BesC this aspartate ring is
replaced by positive lysine residues (K366) and negative aspartate
residues (D363) resulting in a non-selective BesC channel, in
contrast to the cation selective TolC channel. The lysine residue
(K366) and the aspartate residue (D363) could neutralize one
another, meaning that no net charge exists in the tunnel entrance
resulting in a non-selective channel. Another difference between
BesC and TolC is that the BesC channel has a nearly four-fold
higher conductance in 1 M KCl than TolC (80 pS in 1 M KCl)
. TolC of E. coli contains a circular network of inter- and intra-
molecular connections between the three monomers that is
responsible for the stability of the almost closed state of the
channel-tunnel . This network involves residues located at the
inner and outer coiled coil of the tunnel domain, which are not
conserved in BesC (see Figure 4A). This suggests that other
interactions keep the helices of the tunnel entrance together;
otherwise the single-channel conductance would be much higher.
The modeled BesC structure shows that the two charged residues
D363 and K366 of different monomers are in close proximity (2.1
A˚), allowing formation of intermolecular salt bridges. Thus, they
are able to establish a similar circular network keeping the helices
at the tunnel entrance in a close conformation.
The existence of a multi-drug efflux system in B. burgdorferi
makes sense because B. burgdorferi carries natural resistance for
several antibiotics . Resistance can originate from multi-drug
efflux pumps, as has previously been demonstrated for several
Gram-negative bacteria [25,58,59,60].
In this article we described three genes, besB, besC and besA,
which encode proteins comprising a multi-drug efflux system in B.
burgdorferi. We further showed that the BesC protein, a TolC
homolog, is a novel virulence factor of B. burgdorferi. This efflux
system might be part of a Type I secretion machinery for
maintenance of cellular homeostasis or export of exogenous toxic
agents, perhaps necessary for survival in vastly different host
Type 1 Secretion System in Borrelia
PLoS Pathogens | www.plospathogens.org72008 | Volume 4 | Issue 2 | e1000009
environments. The finding of this novel system in borreliae offers
the potential to develop anti-borrelial therapeutics by screening
compounds affecting the efflux system. Further functional studies
will open new possibilities to investigate and understand the
virulence mechanisms of Borrelia spirochetes.
Materials and Methods
Bacterial strains and growth conditions
Infectious, low-passage strain B31  was used for RNA
extraction and operon analysis. Strain p66::str  containing
PflaB-aadA insertion in the p66 gene was used for constructing besC
mutant (p66::str/besC::kan) used in planar lipid bilayer assay. For
animal infectivity studies strain 5A4NP1 , infectious clone of
B31 containing a disruption of nicotinamidase gene pncA (bbe22)
carried on plasmid lp25, was used to create besC knock-out and
complementation strains, respectively, besC::str and besC::str + besC.
Unless otherwise stated, bacteria were grown in BSK-II medium
 supplemented with 6% rabbit serum (Sigma) at 35uC until the
cell density reached approximately 107–108cells ml21. E. coli
Top10 (Invitrogen), used for cloning experiments and E. coli
ROSETTA (Novagen), used for protein overexpression, were
grown at 37uC in Luria-Bertani (LB) broth or on LB agar plates,
containing 50 mg/ml of carbenicillin, 5 mg/ml gentamicin or
50 mg/ml kanamycin when needed.
Sequence analysis and protein modeling
The amino acid sequences of BesC (BB0142), BesA (BB0141)
and BesB (BB0140) were analyzed using BLAST and the
Conserved Domain Database at the National Center for
Biotechnology Information (http://www.ncbi.nlm.nih.gov/). Ho-
mologous sequences were aligned using the ClustalW program
(http://www.ebi.ac.uk/clustalw/). Based on sequence alignments,
protein modeling was performed on the swiss-model platform
(http://swissmodel.expasy.org/; Text S1). For modeling of BesC
(BB0142) the coordinates of OprM (pdb: 1WP1), for BesA
(BB0141) the coordinates of MexA (pdb: 1T5E) and for BesB
(BB0140) the coordinates of AcrB (pdb: 2HRT) were used.
Construction of besC gene inactivation and
Plasmid pOK-besC::kan was constructed to inactivate the besC
gene in B. burgdorferi p66::str strain. Primers besC-XhoI-f and besC-
BamHI-r (Table S1) containing XhoI and BamHI restriction sites
were used to amplify a fragment of 2189 bp, covering the besC
gene, 426 bp upstream and 440 bp downstream from the gene.
Purified PCR product was ligated into BamHI- and XboI-digested
pOK12, a 2.1-kb low-copy-number plasmid containing a
kanamycin resistance gene . By using primers besC-NcoI-f
and besC-PstI-r (Table S1) containing NcoI and PstI restriction sites
the whole plasmid was amplified and a 1.3-kb PflaB-kanamycin
fragment described elsewhere  was amplified with primers
kan-F-PstI and kan-R-NcoI. The kan fragment was then subcloned
into the PstI- and NcoI-digested amplicon of the besC gene and
pOK12 vector. A similar strategy was applied to construct the
plasmid pOK-besC::str which was used to inactivate besC gene in B.
burgdorferi 5A4NP1 strain. Except for selection marker, a 1.3-kb
PflgB-streptomycin fragment  was amplified using primers
aada-F-PstI and aada-R-NcoI and subcloned into the PstI- and
NcoI-digested amplicon of besC gene and pOK12 vector.
Complementation plasmid pCOMP was constructed based on
pBSV2G shuttle vector . Primers besABC-BamHI and
besABC-PstI (Table S1) containing BamHI and PstI restriction
sites were used to amplify a 6299 bp fragment, covering the besC,
besA, besB genes with an additional 425 bp upstream and 457 bp
downstream of the genes. This product was subjected to restriction
enzyme digestion and ligated to BamHI- and PstI-digested
pBSV2G shuttle vector.
Electroporation of B. burgdorferi and screening of
Preparation of competent B. burgdorferi cells and electroporation
was done as described previously [68,69]. Single clones were
obtained, as described elsewhere , in the presence of the
appropriate selective antibiotics: streptomycin (50 mg ml), genta-
mycin (40 mg ml), and/or kanamycin (200 mg ml). Transformants
were further analyzed using primers described in Table S1.
Determination of the plasmid profile and recovery of
Total genomic DNA from B. burgdorferi strains, was prepared
using a Wizard genomic DNA purification kit (Promega). Plasmid
contents of the B. burgdorferi strains 5A4NP1, besC mutant, and
complemented besC mutant were determined by PCR as
previously described by Elias et al. . The B. burgdorferi strains
recovered from the mouse organs were also tested for plasmid
content as described above. To determine the presence of the
complementation plasmid in the complemented besC mutant
recovered from mouse organs, total DNA from this strain was
transformed into E. coli cells. Gentamycin-resistant clones were
screened by PCR for the presence of the shuttle vector.
Separation of the outer membrane proteins of B.
burgdorferi and purification of BesC
Outer membrane proteins (B-fractions) of B. burgdorferi strains
p66::str (a clone with an inactivated p66 gene) and p66::str/besC::kan
(a clone with inactivated besC and p66 genes) were prepared by
detergent (octyl-glucopyranoside) extraction as described else-
where . Purification of native BesC was performed by anion
exchange MonoQ chromatography in combination with fast
protein liquid chromatography (FPLC) (Amersham Biosciences).
About 200 mg of p66::str fraction B was dissolved in 800 ml 2%
lauryl-dimethyl-amine-oxide (LDAO, Sigma) and applied to the
column. The column was first washed with 7.5 ml 0.4% (LDAO)
buffered with 10 mM Tris-HCl (pH 8.0). Bound proteins were
eluted with a linear NaCl gradient (0 to 1 M) containing 0.4%
LDAO buffered with 10 mM Tris-HCl (pH 8.0). Fractions
showing a peak in the FPLC chromatogram were further analyzed
by SDS-PAGE and black lipid bilayer assay. A control FPLC was
performed as described above with about 200 mg of p66::str/
besC::kan fraction B.
Overexpression of recombinant constructs of BesC, BesA,
For overproduction of BesC, BesA, and BesB fragments in E. coli
ROSETTA, the pET-M11 plasmid  was used. The gene
fragments were amplified by PCR by using oligonucleotides
described in Table S1. After restriction enzyme digestion, the PCR
products were ligated into the plasmid pET-M11. The E. coli cells
carrying expression plasmids were grown at 37uC to OD600= 0.6
in LB medium containing 50 mg of kanamycin per ml and protein
expression was induced by addition of isopropyl-b-d-thiogalacto-
pyranoside (IPTG) to a final concentration of 1 mM. The culture
was incubated further for 2 h, and cells were collected by
centrifugation at 6,000 6g for 15 min. The cells were suspended
in 0.1 culture volume of 20 mM Tris-HCl (pH 8.0). Lysozyme
(Sigma) was added to a final concentration of 0.1 mg/ml, and the
Type 1 Secretion System in Borrelia
PLoS Pathogens | www.plospathogens.org8 2008 | Volume 4 | Issue 2 | e1000009
cells were disrupted by sonication. The soluble and insoluble
fractions were separated by centrifugation at 10,000 6 g for
15 min. The recombinant protein was expressed in inclusion
bodies. Inclusion bodies were washed twice with 40 ml of wash
buffer (20 mM Tris-HCl [pH 7.5], 10 mM EDTA, 1% Triton X-
100) and used to raise antiserum against B. burgdorferi BesC, BesA,
Polyclonal antiserum was raised against recombinant fragments
of BesC, BesA, and BesB produced as described above. One
milligram of washed inclusion bodies was separated on a sodium
dodecyl sulfate (SDS)—12.5% polyacrylamide gel electrophoresis
(PAGE) gel. Recombinant protein was excised from the gel and
approximately 100 mg of protein was used for rabbit immuniza-
tion and subsequent boosts. As a control, murine monoclonal
antibody H9724 , which recognizes the constitutively
expressed FlaB (flagellin) protein was used.
Protein electrophoresis, immunoblotting, and antiserum
Total B. burgdorferi proteins were prepared from cells grown to
stationary phase by harvesting the cells by centrifugation, and
washing twice in phosphate-buffered saline. For gel electrophoresis,
proteins were boiled for 5 min in NuPAGE sample buffer and
separated through 4 to 12% NuPAGE bis-Tris polyacrylamide gels
(Invitrogen). For immunoblotting, proteins were transferred to a
polyvinylidene difluoride membrane (PVDF) (PALL Corporation)
and probed with antibodies. Bound antibodies were detected using
peroxidase-conjugated anti-rabbit or anti-mouse antibodies (DAKO
A/S) and enhanced chemiluminescence reagents according to the
manufacturer’s instructions (Amersham Pharmacia Biotech).
Planar lipid bilayer assay
The methods used for the black lipid bilayer experiments have
been described previously . For preparation of the artificial lipid
membranes, a 1% solution of diphytanoyl phosphatidylcholine
(Avanti Polar Lipids) in n-decane was used. All salts (analytical
grade) werepurchasedfrom Merck. Theaqueoussalt solutions were
used without buffering and had a pH around 6 unless otherwise
indicated. The temperature was kept at 20uC throughout. The
channel-forming protein solutions were diluted in 1% Genapol X-
080 (Fluka) and added to the aqueous phase after the membrane
turned black. The membrane current was measured with a pair of
calomel electrodes switched in series with a voltage source and an
electrometer (Keithley 617). For single-channel recordings the
electrometer was replaced by a highly sensitive current amplifier
(Keithley 427). The amplified signal was recorded with a strip chart
recorder. The zero-current membrane potentials were measured as
described previously . The membranes were formed in a
100 mM salt solution containing a predetermined protein concen-
tration so that the membrane conductance increased about 100–
1000 foldwithin10–20 minaftermembrane formation. Atthistime
the instrumentation was switched to the measurements of the zero-
current potential and the salt concentration on one side of the
membrane was raised by adding small amounts of concentrated salt
solutions. The zero-current membrane potential reached its final
value after 10 min. The voltage dependence of the porin channel
wascheckedas described elsewhere  usingmembranepotentials
as high as 2150 to +150 mV.
The FPLC fractions showing channel-forming activity were
subjected to SDS-PAGE followed by silver staining . The
different bands were analyzed by nano LC-mass spectrometry as
described elsewhere . Data interpretation of the MS/MS
datasets was performed by the Mascot algorithm . In detail an
Ultimate II (LC Packings, Germering, Germany) in combination
wereused.Massspectraobtained byLC-MS/MSanalysiswere used
to identify the corresponding peptides with the MascotTM 
(version 2.1.6). The algorithm searched in the NCBI database
(16.08.2005) restricted to the bacteria protein taxonomy with the
following parameter set: (a) fixed modification: carbamidomethyl
(C);(b) variable modification: oxidation (M); (c) peptide and MS/MS
tolerance: +/2 1.5 Da; (d) ion score cut-off: 30.
Microdilution susceptibility testing, determination of MIC
and MBC values
Compounds tested belonged to classes of penicillins (penicillin
G, carbenicillin; Sigma-Aldrich), tetracyclines (tetracycline; Sigma-
Aldrich), macrolides (azithromycin; Sigma-Aldrich), cephalospo-
rins (cefotaxime; Sigma-Aldrich), detergents (sodium dodecyl
sulfate; Scharlau Chemie) and intercalators (ethidium bromide;
Bio-Rad). To test different concentrations, serial dilutions of
substances were done in 96-well Microtest plates (Falcon). A
colorimetric assay was used for susceptibility testing as described
elsewhere [80,81]. Briefly, B. burgdorferi 5A4NP1, besC::str and the
complemented strain were cultured in BSK-II  at 35uC to log
phase and adjusted to 56107bacteria per ml as determined by
enumeration with a Petroff-Hausser bacteria counting chamber
(C.A. Hausser & Son). Microtiter wells were seeded with 56106
bacteria and antibiotic in a final volume of 200 ml. The ranges of
concentrations tested were as follows (mg/ml): penicillin G 0.001–
20; carbenicillin 0.00015–5; tetracycline 0.00035–10, azithromy-
cin 0.00006–2; cefotaxime 0.001–25; ethidium bromide 0.002–50
and sodium dodecyl sulfate 1.5–50000. Microtiter trays with
Borrelia samples and growth controls were covered with a low
evaporation lid supplied by the plate manufacturer (Falcon) and
cultured at 35uC with 1% CO2. Growth was examined after 0, 24,
48, and 72 h by measurement of indicator color shift at 562/
630 nm using an ELISA reader (Multiscan RC, Labsystems) in
combination with a software-assisted calculation program (Genesis
Lite 3.03, Life Sciences Ltd).
Colorimetric MIC’s of isolates were measured in triplicate by
quantification of growth utilizing a software-assisted calculation
(Genesis Lite 3.03, Life Sciences Ltd.). Growth of samples was
determined for each well based on the decrease of absorbance
after 72 h (Et72) in comparison to the initial absorbance values
(Eto). The lowest concentration of antibiotic at which no color shift
could be detected was interpreted as the MIC.
Minimal borreliacidal concentration (MBC) values for tested
compounds were determined in following way. Aliquots (50 ml) from
all vials without demonstrable growth were inoculated into 5 ml of
fresh BSK medium (dilution factor 1:100) to achieve a sample
dilution below the MIC and then incubated at 35uC at 1% CO2for
an additional 3 weeks. After gentle agitation of the subcultures, 5–10
high power fields were examined by dark-field microscopy for the
presence or absence of viable Borrelia. The MBC was defined as the
lowest concentration of the antimicrobial agent where no spirochetes
could be detected after three weeks of subculture.
For each strain and substance three independent experiments
were performed on different days.
Isolation of RNA and RT-PCR
All reagents were prepared with diethylpyrocarbonate (DEPC)-
treated water. Total RNA was isolated from in vitro-cultured
Borrelia using the Ultraspec-II RNA isolation system (Biotex
Type 1 Secretion System in Borrelia
PLoS Pathogens | www.plospathogens.org92008 | Volume 4 | Issue 2 | e1000009
Laboratories) according to the manufacturer’s instructions. RNA
was quantified using a NanoDrop spectrophotometer (NanoDrop
Technologies). To remove contaminating genomic DNA, RNA
samples were treated with 3 units of RNase-free DNaseI (Roche).
For RT-PCR the Superscript One-Step RT-PCR with Platinum
Taq kit (Invitrogen) was used following the manufacturer’s
instructions. A negative control with sterile water was used to
verify the purity of the reagents. The absence of DNA
contamination was verified by PCR. Ten microliters of each
RT-PCR product was analyzed on a 1% agarose gel stained with
ethidium bromide (5 mg ml21). The primers b1 and a1 were used
to detect a transcript spanning besB (bb0140) and besA (bb0141), the
primers a2 and c1 were used to amplify a transcript spanning besA
(bb0141) and besC (bb0142), the primers b1 and c1 were used to
detect a transcript spanning besB (bb0140) and besC (bb0142).
Primer sequences are shown in Table S1.
Animal infectivity studies
B. burgdorferi 5A4NP1 infectious strain, besC::str mutant strain
and besC::str + besC complemented strain were used for mouse
infections. Seven four-week-old C3H/HeN mice (Bomholt Ga ˚rd
Breeding) per strain were subcutaneously injected with 106
spirochetes in 0.1 ml culture medium. The number of bacteria
was determined microscopically in a Petroff–Hausser chamber.
After 2 weeks mice were sacrificed and heart, bladder, knee, and
ear were removed from each mouse and incubated for 4 weeks in
BSK-II medium containing 7% rabbit serum and supplemented
withsulfamethoxazole(1.25 ml ml21)
(4 ml ml21). Each sample was examined for the presence of
spirochetes by dark field microscopy.
The GenBank (http://www.ncbi.nlm.nih.gov/Genbank/index.
html) accession numbers of guinea pig genes used for primer design
are as follows: b-actin (AF508792.1); IFNc (AY151287.1); IL-1b
(AF119622); MCP-1 (L04985); MCP-3 (AB014340); RANTES
(CPU77037); TLR3 (DQ415679.1); and TNFa (CPU77036).
The Protein Data Bank (http://www.rcsb.org/pdb/) ID
numbers for the structures discussed in this paper are crystal form
1 (2Q8A) and crystal form 2 (2Q8B).
Dp66::str were separated by anion exchange chromatography.
Fractions showing uniform channel-forming activity with a single-
channel conductance of about 300 pS were precipitated, separated
by SDS-PAGE and silver stained.
Found at: doi:10.1371/journal.ppat.1000009.s001 (0.07 MB
Outer membrane proteins of B. burgdorferi B31-A
Found at: doi:10.1371/journal.ppat.1000009.s002 (0.06 MB
Oligonucleotide primers used in this study.
Found at: doi:10.1371/journal.ppat.1000009.s003 (0.07 MB
Identification of BesC (BB0142) by peptide mass
Found at: doi:10.1371/journal.ppat.1000009.s004 (0.08 MB
We thank Elke Maier and Bettina Schiffler for the help with the black-lipid
membrane experiments and Pa ¨r Comstedt for help with the animal
experiments. Betty Guo is greatly acknowledged for carefully reading the
Conceived and designed the experiments: IB KD YO SB. Performed the
experiments: IB KD. Analyzed the data: IB KD YO CA RB SB.
Contributed reagents/materials/analysis tools: RB SB. Wrote the paper:
IB YO CA RB SB.
1. Benach JL, Bosler EM, Hanrahan JP, Coleman JL, Habicht GS, et al. (1983)
Spirochetes isolated from the blood of two patients with Lyme disease.
N Engl J Med 308: 740–742.
2. Steere AC, Grodzicki RL, Kornblatt AN, Craft JE, Barbour AG, et al. (1983)
The spirochetal etiology of Lyme disease. N Engl J Med 308: 733–740.
3. Steere AC, Coburn J, Glickstein L (2004) The emergence of Lyme disease. J Clin
Invest 113: 1093–1101.
4. Radolf JD, Goldberg MS, Bourell K, Baker SI, Jones JD, et al. (1995)
Characterization of outer membranes isolated from Borrelia burgdorferi, the
Lyme disease spirochete. Infect Immun 63: 2154–2163.
5. Radolf JD, Bourell KW, Akins DR, Brusca JS, Norgard MV (1994) Analysis of
Borrelia burgdorferi membrane architecture by freeze-fracture electron
microscopy. J Bacteriol 176: 21–31.
6. Walker EM, Borenstein LA, Blanco DR, Miller JN, Lovett MA (1991) Analysis
of outer membrane ultrastructure of pathogenic Treponema and Borrelia species
by freeze-fracture electron microscopy. J Bacteriol 173: 5585–5588.
7. Noppa L, O¨stberg Y, Lavrinovicha M, Bergstro ¨m S (2001) P13, an integral
membrane protein of Borrelia burgdorferi, is C-terminally processed and
contains surface-exposed domains. Infect Immun 69: 3323–3334.
8. O¨stberg Y, Pinne M, Benz R, Rosa P, Bergstro ¨m S (2002) Elimination of
channel-forming activity by insertional inactivation of the p13 gene in Borrelia
burgdorferi. J Bacteriol 184: 6811–6819.
9. Pinne M, Denker K, Nilsson E, Benz R, Bergstro ¨m S (2006) The BBA01 protein,
a member of paralog family 48 from Borrelia burgdorferi, is potentially
interchangeable with the channel-forming protein P13. J Bacteriol 188:
10. Skare JT, Mirzabekov TA, Shang ES, Blanco DR, Erdjument-Bromage H, et al.
(1997) The Oms66 (p66) protein is a Borrelia burgdorferi porin. Infect Immun
11. Nikaido H (1998) Multiple antibiotic resistance and efflux. Curr Opin Microbiol
12. Li XZ, Nikaido H (2004) Efflux-mediated drug resistance in bacteria. Drugs 64:
13. Koronakis V, Sharff A, Koronakis E, Luisi B, Hughes C (2000) Crystal structure
of the bacterial membrane protein TolC central to multidrug efflux and protein
export. Nature 405: 914–919.
14. Andersen C, Hughes C, Koronakis V (2000) Chunnel vision. Export and efflux
through bacterial channel-tunnels. EMBO Rep 1: 313–318.
15. Nikaido H (1994) Prevention of drug access to bacterial targets: permeability
barriers and active efflux. Science 264: 382–388.
16. Nikaido H, Zgurskaya HI (2001) AcrAB and related multidrug efflux pumps of
Escherichia coli. J Mol Microbiol Biotechnol 3: 215–218.
17. Nikaido H (1996) Multidrug efflux pumps of gram-negative bacteria. J Bacteriol
18. Ma D, Alberti M, Lynch C, Nikaido H, Hearst JE (1996) The local repressor
AcrR plays a modulating role in the regulation of acrAB genes of Escherichia
coli by global stress signals. Mol Microbiol 19: 101–112.
19. Zgurskaya HI, Nikaido H (1999) Bypassing the periplasm: reconstitution of the
AcrAB multidrug efflux pump of Escherichia coli. Proc Natl Acad Sci U S A 96:
20. Saier MH Jr, Tam R, Reizer A, Reizer J (1994) Two novel families of bacterial
membrane proteins concerned with nodulation, cell division and transport. Mol
Microbiol 11: 841–847.
21. Fralick JA (1996) Evidence that TolC is required for functioning of the Mar/
AcrAB efflux pump of Escherichia coli. J Bacteriol 178: 5803–5805.
22. Dinh T, Paulsen IT, Saier MH Jr (1994) A family of extracytoplasmic proteins
that allow transport of large molecules across the outer membranes of gram-
negative bacteria. J Bacteriol 176: 3825–3831.
23. Yen MR, Peabody CR, Partovi SM, Zhai Y, Tseng YH, et al. (2002) Protein-
translocating outer membrane porins of Gram-negative bacteria. Biochim
Biophys Acta 1562: 6–31.
24. Marchler-Bauer A, Bryant SH (2004) CD-Search: protein domain annotations
on the fly. Nucleic Acids Res 32: W327–331.
Type 1 Secretion System in Borrelia
PLoS Pathogens | www.plospathogens.org 102008 | Volume 4 | Issue 2 | e1000009
25. Sulavik MC, Houseweart C, Cramer C, Jiwani N, Murgolo N, et al. (2001)
Antibiotic susceptibility profiles of Escherichia coli strains lacking multidrug
efflux pump genes. Antimicrob Agents Chemother 45: 1126–113.
26. Pinne M, Thein M, Denker K, Benz R, Coburn J, et al. (2007) Elimination of
channel-forming activity by insertional inactivation of the p66 gene in Borrelia
burgdorferi. FEMS Microbiol Lett 266: 241–249. 6.
27. Benz R, Janko K, Lauger P (1979) Ionic selectivity of pores formed by the matrix
protein (porin) of Escherichia coli. Biochim Biophys Acta 551: 238–247.
28. Akama H, Kanemaki M, Yoshimura M, Tsukihara T, Kashiwagi T, et al. (2004)
Crystal structure of the drug discharge outer membrane protein, OprM, of
Pseudomonas aeruginosa: dual modes of membrane anchoring and occluded
cavity end. J Biol Chem 279: 52816–52819.
29. Federici L, Du D, Walas F, Matsumura H, Fernandez-Recio J, et al. (2005) The
crystal structure of the outer membrane protein VceC from the bacterial
pathogen Vibrio cholerae at 1.8 A resolution. J Biol Chem 280: 15307–15314.
30. Polleichtner G, Andersen C (2006) The channel-tunnel HI1462 of Haemophilus
31. Stegmeier JF, Polleichtner G, Brandes N, Hotz C, Andersen C (2006)
Importance of the adaptor (membrane fusion) protein hairpin domain for the
functionality of multidrug efflux pumps. Biochemistry 45: 10303–10312.
32. Mursic VP, Wilske B, Schierz G, Holmburger M, Suss E (1987) In vitro and in
vivo susceptibility of Borrelia burgdorferi. Eur J Clin Microbiol 6: 424–426.
33. Hansen K, Hovmark A, Lebech AM, Lebech K, Olsson I, et al. (1992)
Roxithromycin in Lyme borreliosis: discrepant results of an in vitro and in vivo
animal susceptibility study and a clinical trial in patients with erythema migrans.
Acta Derm Venereol 72: 297–300.
34. Hassler D, Zoller L, Haude M, Hufnagel HD, Heinrich F, et al. (1990)
Cefotaxime versus penicillin in the late stage of Lyme disease–prospective,
randomized therapeutic study. Infection 18: 16–20.
35. Preac-Mursic V, Weber K, Pfister HW, Wilske B, Gross B, et al. (1989) Survival
of Borrelia burgdorferi in antibiotically treated patients with Lyme borreliosis.
Infection 17: 355–359.
36. Lomovskaya O, Watkins WJ (2001) Efflux pumps: their role in antibacterial drug
discovery. Curr Med Chem 8: 1699–1711.
37. Gil H, Platz GJ, Forestal CA, Monfett M, Bakshi CS, et al. (2006) Deletion of
TolC orthologs in Francisella tularensis identifies roles in multidrug resistance
and virulence. Proc Natl Acad Sci U S A 103: 12897–12902.
38. Buckley AM, Webber MA, Cooles S, Randall LP, La Ragione RM, et al. (2006)
The AcrAB-TolC efflux system of Salmonella enterica serovar Typhimurium
plays a role in pathogenesis. Cell Microbiol 8: 847–856.
39. Hirakata Y, Srikumar R, Poole K, Gotoh N, Suematsu T, et al. (2002)
Multidrug efflux systems play an important role in the invasiveness of
Pseudomonas aeruginosa. J Exp Med 196: 109–118.
40. Jerse AE, Sharma ND, Simms AN, Crow ET, Snyder LA, et al. (2003) A
gonococcal efflux pump system enhances bacterial survival in a female mouse
model of genital tract infection. Infect Immun 71: 5576–5582.
41. Nishino K, Latifi T, Groisman EA (2006) Virulence and drug resistance roles of
multidrug efflux systems of Salmonella enterica serovar Typhimurium. Mol
Microbiol 59: 126–141.
42. Lee EH, Shafer WM (1999) The farAB-encoded efflux pump mediates resistance
of gonococci to long-chained antibacterial fatty acids. Mol Microbiol 33:
43. Lacroix FJ, Cloeckaert A, Grepinet O, Pinault C, Popoff MY, et al. (1996)
Salmonella typhimurium acrB-like gene: identification and role in resistance to
biliary salts and detergents and in murine infection. FEMS Microbiol Lett 135:
44. Poole K, Krebes K, McNally C, Neshat S (1993) Multiple antibiotic resistance in
Pseudomonas aeruginosa: evidence for involvement of an efflux operon.
J Bacteriol 175: 7363–7372.
45. Koronakis V (2003) TolC–the bacterial exit duct for proteins and drugs. FEBS
Lett 555: 66–71.
46. Andersen C (2003) Channel-tunnels: outer membrane components of type I
secretion systems and multidrug efflux pumps of Gram-negative bacteria. Rev
Physiol Biochem Pharmacol 147: 122–165.
47. Lobedanz S, Bokma E, Symmons MF, Koronakis E, Hughes C, et al. (2007) A
periplasmic coiled-coil interface underlying TolC recruitment and the assembly
of bacterial drug efflux pumps. Proc Natl Acad Sci U S A 104: 4612–4617.
48. Tamura N, Murakami S, Oyama Y, Ishiguro M, Yamaguchi A (2005) Direct
interaction ofmultidrugeffluxtransporterAcrBandoutermembranechannel TolC
detected via site-directed disulfide cross-linking. Biochemistry 44: 11115–11121.
49. Johnson JM, Church GM (1999) Alignment and structure prediction of
divergent protein families: periplasmic and outer membrane proteins of bacterial
efflux pumps. J Mol Biol 287: 695–715.
50. Gupta RS (2000) The natural evolutionary relationships among prokaryotes.
Crit Rev Microbiol 26: 111–131.
51. Sickmann A, Mreyen M, Meyer HE (2002) Identification of modified proteins
by mass spectrometry. IUBMB Life 54: 51–57.
52. Benz R, Maier E, Gentschev I (1993) TolC of Escherichia coli functions as an
outer membrane channel. Zentralbl Bakteriol 278: 187–196.
53. Andersen C (2004) Drug efflux and protein export through channel-tunnels;
Benz R, ed. Weinheim, Germany: WILEY-VCH Verlag GmbH. pp 139–167.
54. Andersen C, Hughes C, Koronakis V (2001) Protein export and drug efflux
through bacterial channel-tunnels. Curr Opin Cell Biol 13: 412–416.
55. Andersen C, Koronakis E, Hughes C, Koronakis V (2002) An aspartate ring at
the TolC tunnel entrance determines ion selectivity and presents a target for
blocking by large cations. Mol Microbiol 44: 1131–1139.
56. Andersen C, Koronakis E, Bokma E, Eswaran J, Humphreys D, et al. (2002)
Transition to the open state of the TolC periplasmic tunnel entrance. Proc Natl
Acad Sci U S A 99: 11103–11108.
57. Hunfeld KP, Weigand J, Wichelhaus TA, Kekoukh E, Kraiczy P, et al. (2001) In
vitro activity of mezlocillin, meropenem, aztreonam, vancomycin, teicoplanin,
ribostamycin and fusidic acid against Borrelia burgdorferi. Int J Antimicrob
Agents 17: 203–208.
58. Mata MT, Baquero F, Perez-Diaz JC (2000) A multidrug efflux transporter in
Listeria monocytogenes. FEMS Microbiol Lett 187: 185–188.
59. Eswaran J, Koronakis E, Higgins MK, Hughes C, Koronakis V (2004) Three’s
company: component structures bring a closer view of tripartite drug efflux
pumps. Curr Opin Struct Biol 14: 741–747.
60. Grkovic S, Brown MH, Skurray RA (2002) Regulation of bacterial drug export
systems. Microbiol Mol Biol Rev 66: 671–701, table of contents.
61. Burgdorfer W, Barbour AG, Hayes SF, Benach JL, Grunwaldt E, et al. (1982)
Lyme disease-a tick-borne spirochetosis? Science 216: 1317–1319.
62. Kawabata H, Norris SJ, Watanabe H (2004) BBE02 disruption mutants of
Borrelia burgdorferi B31 have a highly transformable, infectious phenotype.
Infect Immun 72: 7147–7154.
63. Barbour AG (1984) Isolation and cultivation of Lyme disease spirochetes.
Yale J Biol Med 57: 521–525.
64. Vieira J, Messing J (1991) New pUC-derived cloning vectors with different
selectable markers and DNA replication origins. Gene 100: 189–194.
65. Bono JL, Elias AF, Kupko JJ 3rd, Stevenson B, Tilly K, et al. (2000) Efficient
targeted mutagenesis in Borrelia burgdorferi. J Bacteriol 182: 2445–2452.
66. Frank KL, Bundle SF, Kresge ME, Eggers CH, Samuels DS (2003) aadA confers
streptomycin resistance in Borrelia burgdorferi. J Bacteriol 185: 6723–6727.
67. Elias AF, Bono JL, Kupko JJ 3rd, Stewart PE, Krum JG, et al. (2003) New
antibiotic resistance cassettes suitable for genetic studies in Borrelia burgdorferi.
J Mol Microbiol Biotechnol 6: 29–40.
68. Samuels DS, Mach KE, Garon CF (1994) Genetic transformation of the Lyme
disease agent Borrelia burgdorferi with coumarin-resistant gyrB. J Bacteriol 176:
69. Tilly K, Elias AF, Bono JL, Stewart P, Rosa P (2000) DNA exchange and
insertional inactivation in spirochetes. J Mol Microbiol Biotechnol 2: 433–442.
70. Yang XF, Pal U, Alani SM, Fikrig E, Norgard MV (2004) Essential role for
OspA/B in the life cycle of the Lyme disease spirochete. J Exp Med 199:
71. Elias AF, Stewart PE, Grimm D, Caimano MJ, Eggers CH, et al. (2002) Clonal
polymorphism of Borrelia burgdorferi strain B31 MI: implications for
mutagenesis in an infectious strain background. Infect Immun 70: 2139–2150.
72. Magnarelli LA, Anderson JF, Barbour AG (1989) Enzyme-linked immunosor-
bent assays for Lyme disease: reactivity of subunits of Borrelia burgdorferi.
J Infect Dis 159: 43–49.
73. Pinotsis N, Petoukhov M, Lange S, Svergun D, Zou P, et al. (2006) Evidence for
a dimeric assembly of two titin/telethonin complexes induced by the telethonin
C-terminus. Journal of Structural Biology 155: 239–250.
74. Barbour AG, Hayes SF, Heiland RA, Schrumpf ME, Tessier SL (1986) A
Borrelia-specific monoclonal antibody binds to a flagellar epitope. Infect Immun
75. Benz R, Janko K, Boos W, Lauger P (1978) Formation of large, ion-permeable
membrane channels by the matrix protein (porin) of Escherichia coli. Biochim
Biophys Acta 511: 305–319.
76. Riess FG, Benz R (2000) Discovery of a novel channel-forming protein in the cell
wall of the non-pathogenic Nocardia corynebacteroides. Biochim Biophys Acta
77. Helmut Blum HBHJG (1987) Improved silver staining of plant proteins, RNA
and DNA in polyacrylamide gels. Electrophoresis 8: 93–99.
78. Sickmann A, Reinders J, Wagner Y, Joppich C, Zahedi R, et al. (2003) The
proteome of Saccharomyces cerevisiae mitochondria. Proc Natl Acad Sci U S A
79. Perkins DN, Pappin DJ, Creasy DM, Cottrell JS (1999) Probability-based
protein identification by searching sequence databases using mass spectrometry
data. Electrophoresis 20: 3551–3567.
80. Hunfeld KP, Kraiczy P, Wichelhaus TA, Schafer V, Brade V (2000)
Colorimetric in vitro susceptibility testing of penicillins, cephalosporins,
macrolides, streptogramins, tetracyclines, and aminoglycosides against Borrelia
burgdorferi isolates. Int J Antimicrob Agents 15: 11–17.
81. Hunfeld KP, Kraiczy P, Wichelhaus TA, Schafer V, Brade V (2000) New
colorimetric microdilution method for in vitro susceptibility testing of Borrelia
burgdorferi against antimicrobial substances. Eur J Clin Microbiol Infect Dis 19:
82. Higgins MK, Bokma E, Koronakis E, Hughes C, Koronakis V (2004) Structure
of the periplasmic component of a bacterial drug efflux pump. Proc Natl Acad
Sci U S A 101: 9994–9999.
83. Seeger MA, Schiefner A, Eicher T, Verrey F, Diederichs K, et al. (2006)
Structural asymmetry of AcrB trimer suggests a peristaltic pump mechanism.
Science 313: 1295–1298.
84. Benz R, Schmid A, Hancock RE (1985) Ion selectivity of Gram-negative
bacterial porins. J Bacteriol 162: 722–727.
Type 1 Secretion System in Borrelia
PLoS Pathogens | www.plospathogens.org 11 2008 | Volume 4 | Issue 2 | e1000009