A conservative amino acid change alters the function of BosR, the redox regulator of Borrelia burgdorferi

Article (PDF Available)inMolecular Microbiology 54(5):1352-63 · January 2005with19 Reads
DOI: 10.1111/j.1365-2958.2004.04352.x · Source: PubMed
  • 29.41 · University of Texas at San Antonio
  • 19.97 · The Tauri Group
  • 18.9 · Texas A&M University System Health Science Center
  • 31.74 · Texas A&M University
Abstract
Borrelia burgdorferi, the aetiologic agent of Lyme disease, modulates gene expression in response to changes imposed by its arthropod vector and mammalian hosts. As reactive oxygen species (ROS) are known to vary in these environments, we asked how B. burgdorferi responds to oxidative stress. The B. burgdorferi genome encodes a PerR homologue (recently designated BosR) that represses the oxidative stress response in other bacteria, suggesting a similar function in B. burgdorferi. When we tested the sensitivity of B. burgdorferi to ROS, one clonal non-infectious B. burgdorferi isolate exhibited hypersensitivity to t-butyl hydroperoxide when compared with infectious B. burgdorferi and other non-infectious isolates. Sequence analysis indicated that the hypersensitive non-infectious isolates bosR allele contained a single nucleotide substitution, converting an arginine to a lysine (bosRR39K). Mutants in bosRR39K exhibited an increase in resistance to oxidative stressors when compared with the parental non-infectious strain, suggesting that BosRR39K functioned as a repressor. Complementation with bosRR39K and bosR resulted in differential sensitivity to t-butyl hydroperoxide, indicating that these alleles are functionally distinct. In contrast to BosR, BosRR39K did not activate transcription of a napA promoter-lacZ reporter in Escherichia coli nor bind the napA promoter/operator domain. However, we found that both BosR and BosRR39K bound to the putative promoter/operator region of superoxide dismutase (sodA). In addition, we determined that cells lacking BosRR39K synthesized fourfold greater levels of the decorin binding adhesin DbpA suggesting that BosRR39K regulates genes unrelated to oxidative stress. Based on these data, we propose that the single amino acid substitution, R39K, dramatically alters the activity of BosR by altering its ability to bind DNA at target regulatory sequences.
Molecular Microbiology (2004)
54
(5), 13521363 doi:10.1111/j.1365-2958.2004.04352.x
© 2004 Blackwell Publishing Ltd
Blackwell Science, LtdOxford, UKMMIMolecular Microbiology0950-382XBlackwell Publishing Ltd, 2004
? 2004
54
513521363
Original Article
Redox regulation in B. burgdorferiJ. Seshu
et al.
Accepted 11 August, 2004. *For correspondence. E-mail
jskare@tamu.edu; Tel. (
+
1) 979 845 1376; Fax (
+
1) 979 845 3479.
A conservative amino acid change alters the function of
BosR, the redox regulator of
Borrelia burgdorferi
J. Seshu,
1
Julie A. Boylan,
2
Jenny A. Hyde,
1
Kristen L. Swingle,
1
Frank C. Gherardini
2
and
Jonathan T. Skare
1
*
1
Department of Medical Microbiology and Immunology,
407 Reynolds Medical Building, Texas A&M University
Health Science Center, College Station, TX 77843, USA.
2
Laboratory of Human Bacterial Pathogenesis, Rocky
Mountain Laboratory, NIAID, NIH, Hamilton, MT 59840,
USA.
Summary
Borrelia burgdorferi
, the aetiologic agent of Lyme
disease, modulates gene expression in response to
changes imposed by its arthropod vector and mam-
malian hosts. As reactive oxygen species (ROS) are
known to vary in these environments, we asked how
B. burgdorferi
responds to oxidative stress. The
B.
burgdorferi
genome encodes a PerR homologue
(recently designated BosR) that represses the oxida-
tive stress response in other bacteria, suggesting a
similar function in
B. burgdorferi
. When we tested
the sensitivity of
B. burgdorferi
to ROS, one clonal
non-infectious
B. burgdorferi
isolate exhibited
hypersensitivity to
t
-butyl hydroperoxide when
compared with infectious
B. burgdorferi
and other
non-infectious isolates. Sequence analysis indicated
that the hypersensitive non-infectious isolates
bosR
allele contained a single nucleotide substitution,
converting an arginine to a lysine (
bosRR39K
).
Mutants in
bosRR39K
exhibited an increase in resis-
tance to oxidative stressors when compared with
the parental non-infectious strain, suggesting that
BosRR39K functioned as a repressor. Complemen-
tation with
bosRR39K
and
bosR
resulted in differen-
tial sensitivity to
t
-butyl hydroperoxide, indicating
that these alleles are functionally distinct. In con-
trast to BosR, BosRR39K did not activate trans-
cription of a
napA
promoter–
lacZ
reporter in
Escherichia coli
nor bind the
napA
promoter/opera-
tor domain. However, we found that both BosR and
BosRR39K bound to the putative promoter/operator
region of superoxide dismutase (
sodA
). In addition,
we determined that cells lacking BosRR39K
synthesized fourfold greater levels of the decorin
binding adhesin DbpA suggesting that BosRR39K
regulates genes unrelated to oxidative stress. Based
on these data, we propose that the single amino
acid substitution, R39K, dramatically alters the
activity of BosR by altering its ability to bind DNA at
target regulatory sequences.
Introduction
Experimental data suggests that the tick-borne Lyme
disease agent
Borrelia burgdorferi
modulates gene
expression in both the
Ixodes
vector and mammalian
hosts (Das
et al
., 1997; Ohnishi
et al
., 2001). Several
investigators have demonstrated that gene expression
in
B. burgdorferi
is affected by temperature, pH and
uncharacterized components of the tick bloodmeal
(Schwan
et al
., 1995; Carroll
et al
., 1999; Yang
et al
.,
2000; Ojaimi
et al
., 2003). Subsequent studies showed
that
B. burgdorferi
implanted in the peritoneal cavity of
rats significantly alter gene expression relative to cells
cultivated in BSK-II medium (Akins
et al
., 1998; Revel
et al
., 2002; Brooks
et al
., 2003). More recently, levels
of dissolved oxygen have been shown to alter gene
expression in
B. burgdorferi
(Seshu
et al
., 2004). While
the global effects of these ‘environmental signals’ on
gene expression have been known for some time, the
regulatory mechanisms underlying these changes is
only beginning to be understood. Recently, the regula-
tion of outer surface protein C (
ospC
) and other lipopro-
teins were shown to be regulated by RpoS (
s
S
) which in
turn is regulated by RpoN (
s
54
) (Hübner
et al
., 2001)
and the response regulator Rrp2 (Yang
et al
., 2003).
Within the
B. burgdorferi
genome annotated by TIGR, a
fur
homologue was also identified (Fraser
et al
., 1997).
The
B. burgdorferi fur
locus is of particular interest as
this spirochete contains no enzymes that require iron as
a cofactor (Posey and Gherardini, 2000). The chromo-
somal open reading frame (ORF) BB0647 encoding the
fur
homologue shares significant homology with a regu-
lator of the oxidative stress response designated PerR
(27–35% identical and 50–53% similar depending on
homologue), suggesting that
B. burgdorferi
encodes a
Redox regulation in
B. burgdorferi 1353
© 2004 Blackwell Publishing Ltd,
Molecular Microbiology
,
54
, 13521363
protein with a similar function (Boylan
et al
., 2003). In
Bacillus
spp., PerR has been shown to act as a global
metalloregulatory protein that binds to the operator
region of target genes and represses their expression;
exposure to reactive oxygen species (ROS) results in
reduced binding and derepression of the genes required
to neutralize the oxidative stress response (Herbig and
Helmann, 2001; Fuangthong
et al
., 2002). Absolutely
conserved in all PerR regulators are two separate C–X–
X–C motifs that are believed to co-ordinate divalent met-
als and/or function as a direct redox sensor. Recently,
Boylan
et al
. (2003) demonstrated that the borrelial PerR
homologue functioned as a Zn-dependent transcriptional
activator of ORF BB0690 (designated as a Dps/Dpr
homology or neutrophil activating protein, NapA). Based
on the novel activity of the borrelial PerR homologue,
this regulatory locus was renamed BosR for
Borrelia
oxi-
dative stress regulator (Boylan
et al
., 2003). The BosR-
dependent activation of transcription from the
napA
pro-
moter/operator (P/O)–
lacZ
reporter system in
Escheri-
chia coli
was stimulated by alkyl peroxides suggesting
that BosR functions as a redox regulatory protein (Boy-
lan
et al
., 2003). However, its role as a global regulator
has yet to be demonstrated experimentally. Character-
ization of the BosR regulon would help to define the
genes required for resistance to ROS as the
B. burgdor-
feri
genome predicts only a subset of genes involved in
the oxidative stress response. For example, whereas
B.
burgdorferi
encode a superoxide dismutase homologue
(
sodA
), other important enzymes such as catalase and
alkyl peroxidases are apparently absent.
We have taken a genetic approach to begin to analyse
the regulatory effects of BosR in the presence of ROS.
While numerous attempts to disrupt
bosR
within infec-
tious, low-passage (LP)
B. burgdorferi
isolates have been
unsuccessful,
bosR
was disrupted in a non-infectious,
high-passage (HP) strain B31 derivative specific to our
laboratory. Sequence analysis of this HP
bosR
allele
revealed a single amino acid substitution, which changes
an arginine to a lysine at position 39 (designated
bosRR39K
). Comparative biochemical and
bosRR39K
genetic complementation analyses with wild-type
bosR
and the
bosRR39K
alleles suggest that the R39K amino
acid change in BosR has significantly altered the activity
of this regulatory protein. Interestingly, BosRR39K is
unable to activate transcription of
napA
and bind a DNA
fragment containing the
napA
P/O while retaining its abil-
ity to bind to a putative P/O fragment of superoxide dimu-
tase (
sodA
). Along these lines, we demonstrate herein
that BosRR39K is capable of binding to putative regula-
tory domains of
sodA
and present additional data support-
ing the hypothesis that BosRR39K represses this locus.
These results imply that BosR is a multifunctional protein
with both activator and repressor functions.
Results
A clonal high-passage
B. burgdorferi
strain B31 is
sensitive to
t
-butyl hydroperoxide
The production of ROS is an important host defence
mechanism mediated in response to infection by bacterial
pathogens. Not surprisingly, pathogens have evolved
numerous strategies to protect themselves against the
damaging effects of these agents. One important aspect
of this ROS response is the ability to detoxify a peroxide-
containing environment enough to make it habitable. To
test the ability of
B. burgdorferi
to grow in the presence of
ROS, clones of LP infectious and HP non-infectious
B.
burgdorferi
strain B31 were assayed for their sensitivity to
t
-butyl hydroperoxide (Fig. 1). Our HP isolate is missing
seven of the 21 known plasmids including the 25 kb linear
plasmid (lp25) and one of the 28 kb plasmids (lp28-1) that
are required for colonization and persistence respectively
(Purser and Norris, 2000; Labandeira-Rey and Skare,
2001; Labandeira-Rey
et al
., 2003; Purser
et al
., 2003).
As such, this HP strain is non-infectious in the mouse
animal model system whereas our LP isolate contains all
of the known plasmids and is fully virulent (McDowell
et al
., 2001; Labandeira-Rey
et al
., 2003). To ascertain
whether our HP and LP isolates demonstrated differential
sensitivity to ROS, we incubated these strains with various
oxidative stressors. The LP
B. burgdorferi
strain B31
was resistant to concentrations of
t
-butyl hydroperoxide
greater than 3 mM (Fig. 1). In contrast, our clonal isolate
of HP
B. burgdorferi
strain B31 exhibited an increased
sensitivity to
t
-butyl hydroperoxide, losing nearly five
orders of magnitude in viability in the presence of 1 mM
t
-butyl hydroperoxide relative to the LP isolate. At concen-
trations greater than 1 mM
t
-butyl peroxide, LP
B. burg-
dorferi
cells were still viable whereas our HP
B. burgdorferi
isolate was completely killed indicating a strong differen-
Fig. 1.
Infectious LP
B. burgdorferi
(dark bars) and non-infectious HP
B. burgdorferi
(white bars) show differential sensitivity to
t
-butyl hydro-
peroxide. Data shown are the average of three independent assays
with the bars indicating the standard deviation. Note that HP
B.
burgdorferi
is completely killed by 2 mM and 3 mM
t
-butyl hydroper-
oxide whereas LP
B. burgdorferi
is resistant.
1354
J. Seshu
et al.
© 2004 Blackwell Publishing Ltd,
Molecular Microbiology
,
54
, 13521363
tial in the ability of these isolates to combat oxidative
stressors. It should be noted that other clonal isolates of
HP
B. burgdorferi
strain B31 demonstrate resistance to
ROS comparable to the LP
B. burgdorferi assayed here
(data not shown) indicating that this particular HP B. burg-
dorferi strain B31 isolate is unique in its sensitivity to ROS.
HP BosR contains an arginine to lysine
substitution at residue 39
The inability to respond appropriately to ROS most prob-
ably results from (i) secondary mutations in an oxidative
stress enzyme(s) or (ii) a mutation in a regulator of the
oxidative stress response. Recently, Boylan et al. (2003)
demonstrated that bosR (designated BB0647 by TIGR)
activates expression of the borrelial oxidative stress gene
napA when the napA P/O is fused to the lacZ reporter
gene in E. coli, implying that the sensitivity to ROS
observed in our HP B. burgdorferi isolate may result from
a change in the bosR locus that alters its regulatory activ-
ity. To assess this possibility, DNA was isolated from both
HP and LP B. burgdorferi strain B31 cells. Comparison of
the nucleotide sequence indicated that a single point
mutation was present in our HP bosR isolate, converting
residue 39 from an arginine in LP bosR to a lysine (data
not shown). This allele of bosR is referred to as bosRR39K
for the remainder of this report.
Inactivation of the HP bosRR39K results in increased
resistance to oxidizing agents
In order to assess the role of BosR in oxidative stress
response, we insertionally inactivated the HP B. burgdor-
feri bosRR39K allele using in vitro transposition with a
customized transposon containing a kanamycin resis-
tance locus (whose expression was dependent on the B.
burgdorferi flgB promoter and similar to that reported by
Bono et al., 2000) coupled with allelic exchange. After in
vitro transposition and electroporation into E. coli cells,
two putative transposon mutants of bosRR39K were
tested by polymerase chain reaction (PCR) to confirm the
insertion within this locus. Both candidates yielded the
appropriately sized PCR product and the location of the
custom transposon insertions in bosRR39K was deter-
mined by dideoxy sequencing (data not shown).
The inactivated bosRR39K::kan
R
construct was elec-
troporated into HP B. burgdorferi strain B31 cells. Kana-
mycin-resistant transformants were obtained at a
frequency of 3 ¥ 10
-8
per mg of DNA. Of the 10 transfor-
mants tested by PCR, nine contained the transposon
insertion in the bosRR39K allele (data not shown). Two of
these transformants were analysed by Southern blot
(Fig. 2) and Western immunoblot (data not shown), con-
firming the HP bosRR39K::kan
R
genetic knockout. Both of
the clones analysed contained the desired mutation and
one, designated JS167 (i.e. bosRR39K::kan
R
), was cho-
sen for further studies. To date, we have been unable to
obtain a bosR knockout in LP B. burgdorferi strain B31
nor other isolates of non-infectious B. burgdorferi. It is
important to note that the ability to isolate a bosRR39K
knockout in the HP B. burgdorferi strain used here implies
that additional mutations are present or were selected for
coincident with the bosRR39K genetic inactivation. The
exact nature of these mutations is not known but they, in
addition to the bosRR39K allele and perhaps the absence
of several plasmids known to be missing from the non-
infectious isolate, undoubtedly contribute to the unique
oxidative stress phenotype of the HP B. burgdorferi isolate
used in this study.
The B. burgdorferi strain JS167 and HP parental strain
B31 were tested for their sensitivity to methyl viologen, a
compound that promotes the formation of endogenous
superoxide anion. The bosRR39K::kan
R
strain JS167 was
125-fold more resistant to methyl viologen when com-
pared with the HP strain B31 parent (Fig. 3A) suggesting
that the B. burgdorferi superoxide dismutase (SodA) was
either directly or indirectly derepressed when BosRR39K
is missing. To assess this possibility, cell lysates from both
B. burgdorferi strain JS167 and HP parental strain B31
Fig. 2. Southern blot confirms the mutation in the HP B. burgdorferi
bosRR39K allele.
A and B. DNA from the HP B. burgdorferi parent strain (lanes 1 and
2) and the bosRR39K::kan
R
mutant JS167 (lanes 3 and 4) was
digested with EcoRI (lanes 1 and 3) and EcoRV (lanes 2 and 4) and
probed with (A) the bosR (BB0647)-containing chromosomal region
(indicated by the vertical dotted lines in C) or (B) the kan
R
marker.
The numbers on the left of each panel indicate the size of markers
in kb.
C. A schematic of the bosR chromosomal region with transposition
sites indicated by circular flags showing the customized transposon
insertion (P
flgB
-kan
R
). The numbers show the genomic assignment
within the bosR (BB0647) region of the B. burgdorferi chromosome.
The black and hatched boxes indicate the location of the Tn7 and Tn5
repeats respectively. B, BamHI; RI, EcoRI; RV, EcoRV; S, SalI.
Redox regulation in B. burgdorferi 1355
© 2004 Blackwell Publishing Ltd, Molecular Microbiology, 54, 13521363
were separated on a 10% native acrylamide gel and
assayed for superoxide dismutase activity. Achromatic
zones, corresponding to the location of superoxide dismu-
tase activity, were observed after direct exposure to visible
light. The results obtained indicate that bosRR39K::kan
R
strain (JS167) synthesizes greater levels of SodA as com-
pared with its parental HP strain B31 (Fig. 3B, bottom). A
second faint upper band is also seen that may represent
a interaction between SodA and an unknown borrelial
protein or an additional borrelial protein with superoxide
dismutase-like activity with no homology to Sod proteins
currently in the database (Fig. 3B, top). In addition, we
observed a superoxide dismutase-like background activity
associated with the rabbit serum added to BSK-II medium
(as indicated by an asterisk in Fig. 3B) when diluted sam-
ples of rabbit serum were compared with borrelial protein
lysates (data not shown).
To examine the sensitivity of B. burgdorferi strain JS167
and the HP parental strain B31 to oxidative stressors, we
exposed both isolates to H
2
O
2
and t-butyl hydroperoxide.
JS167 showed significant hyper-resistance to H
2
O
2
(Fig. 4). Specifically, at concentrations of 1 mM and 2 mM
H
2
O
2
, JS167 was 1000-fold and 3140-fold more resistant
to H
2
O
2
relative to its HP isogenic parent strain B31
respectively. Similarly, the bosRR39K::kan
R
strain (JS167)
was dramatically more resistant to t-butyl hydroperoxide
when compared with the HP strain B31 (Fig. 5A). Under
these conditions, strain JS167 was completely unaffected
at concentrations of t-butyl hydroperoxide greater than
3 mM whereas killing of our parent HP strain B31 was
nearly 5 logs greater than JS167 at a concentration of
1 mM. Concentrations of t-butyl hydroperoxide greater
than 1 mM resulted in complete killing of the HP
parent strain. Taken together, the resistance of the
bosRR39K::kan
R
strain to reactive oxygen stressors indi-
cates that BosR regulates the oxidative stress response
in B. burgdorferi and implies that BosRR39K functions as
a repressor in our HP B. burgdorferi isolate.
Genetic complementation with HP and LP bosR alleles
yields distinct phenotypes
To genetically complement the HP B. burgdorferi
bosRR39K::kan
R
strain, JS167 was transformed with
either the B. burgdorferi shuttle vector pKFSS1 alone or
constructs containing either LP or HP alleles of bosR
(designated bosR and bosRR39K respectively). To deter-
mine whether the sensitivity to ROS was restored in the
complemented isolate, we exposed all strains to t-butyl
hydroperoxide as described above. The results indicated
that the bosRR39K::kan
R
strain JS167 was effectively
complemented by the HP bosRR39K allele but only par-
tially by the LP bosR allele (Fig. 5A). As only partial com-
plementation with LP wild-type bosR was attained, we
confirmed by Western immunoblotting that plasmid-
Fig. 3. The bosRR39K::kan
R
knockout mutant is more resistant to
methyl viologen and produces more superoxide dismutase (SodA).
A. The bosRR39K::kan
R
strain (dark bars) and the HP B. burgdorferi
parent (bosRR39K, white bars) were assayed for their sensitivity to
methyl viologen. Data shown are the average of three independent
assays with the bars indicating the standard deviation. The asterisk
(*) denotes a significant difference in methyl viologen sensitivity
(P < 0.05).
B. Superoxide dismutase (SodA) activity is greater in the
bosRR39K::kan
R
strain. B. burgdorferi cells were grown under
microaerophilic conditions and equivalent amounts of protein from HP
B. burgdorferi (lane 1) and JS167 (bosRR39K::kan
R
; lane 2) were
resolved on a native polyacrylamide gel and assayed for SodA activ-
ity. Arrows indicate the location of the SodA activity relative to back-
ground associated with rabbit serum used in the BSK-II growth
medium (indicated by an asterisk).
Concentration of methyl viologen (mM)
Dilution
factor
(log
10
)
supporting
growth
Fig. 4. The bosRR39K::kan
R
knockout mutant is more resistant to
H
2
O
2
. The bosRR39K::kan
R
mutant (dark bars) and HP B. burgdorferi
parent (white bars) were assayed for their sensitivity to H
2
O
2
. Data
shown are the average of three independent assays with the bars
indicating the standard deviation. The asterisk (*) and double asterisk
(**) denote a significant difference in H
2
O
2
sensitivity (P = 0.003 and
P = 0.0004 respectively).
Dilution
factor
(log
10
)
supporting
growth
1356 J. Seshu et al.
© 2004 Blackwell Publishing Ltd, Molecular Microbiology, 54, 13521363
encoded LP BosR and our HP BosRR39K were being
synthesized. Expression of complemented bosRR39K
(Fig. 5B, lane 3) was greater than both the chromosomally
encoded HP parental strain and complemented LP bosR
(Fig. 5B, lanes 1 and 4 respectively), however, the degree
of t-butyl hydroperoxide resistance for the HP bosRR39K
sample was not greater than that observed for the HP
parental strain (Fig. 5A) indicating that the increased
expression of HP bosRR39K in the JS167/pJS239 back-
ground did not alter the physiologic response to ROS.
Furthermore, the immunoblot analysis indicates that
equivalent levels of BosR are synthesized in the parental
HP strain B31 and the bosRR39K knockout strain JS167
complemented with the LP wild-type bosR allele, yet these
two strains exhibit dramatic differences in their sensitivity
to t-butyl hydroperoxide (Fig. 5A). Taken together, these
results imply that the differential sensitivity to t-butyl hydro-
peroxide does not result from the levels of BosR or
BosRR39K observed and suggest that the HP and LP
forms of BosR are functionally distinct.
Another notable aspect of the genetic complementation
analysis is the inability of the LP bosR allele to restore
resistance to t-butyl peroxide in the bosRR39K::kan
R
strain to the levels seen in our hyper-resistant LP infec-
tious B. burgdorferi isolate (Fig. 1). The differential in sen-
sitivity observed reiterates our prior contention that
additional secondary mutations are present within our
hypersensitive HP B. burgdorferi isolate in addition to the
bosRR39K allele. However, our HP B. burgdorferi isolate
lacks several plasmids present in LP B. burgdorferi that
could play a role in the increased sensitivity observed
although no genes with significant similarity to those
involved in the oxidative stress response map to any of
these plasmids. Therefore, assuming that the genetic
information common to our HP isolate and the LP strain
explains the sensitivity to oxidative stressors observed,
then one would predict that the hyper-resistant phenotype
observed in the LP B. burgdorferi isolate would be mir-
rored in the bosRR39K::kan
R
isolate complemented with
LP bosR. Despite this disparity, we contend that the phe-
notype of the bosRR39K::kan
R
mutant relative to its
isogenic parent and LP B. burgdorferi provides a strong
foundation to decipher differences in the oxidative stress
response and biochemical properties as they relate to
both BosR and BosRR39K regulatory activities.
An additional curious feature of the BosRR39K protein
is its anomalous migration relative to BosR after SDS-
PAGE (Fig. 5B, lanes 1 and 3 relative to lane 4). Inasmuch
as the R39K substitution is conservative, it is unclear as
to why these two forms of BosR would resolve so differ-
ently. One possibility is that the mutant BosRR39K protein
folds anomalously and that this effect is not completely
relieved by denaturation in SDS. Although it is also pos-
sible that a post-translational modification might be
responsible for the differential migration observed for
BosRR39K relative to BosR, this seems unlikely as the
same migration pattern is seen when recombinant forms
of these proteins are produced in E. coli cells (data not
shown).
BosR and BosRR39K are functionally distinct
To address the functional differences between BosR and
BosRR39K, we tested their ability to regulate expression
of the B. burgdorferi napA promoter linked to the lacZ
reporter gene in E. coli. Without induction of the LP bosR
Fig. 5. The HP B. burgdorferi bosRR39K::kan
R
mutant is more resistant to t-butyl hydroperoxide; complementation with the bosRR39K allele
restores sensitivity.
A. The parent strain (intact bosRR39K; white columns), bosRR39K::kan
R
mutant (hatched columns), bosRR39K::kan
R
mutant with the shuttle
vector pKFSS1 (dark columns), pKFSS1 containing LP bosR (pJS232; bricked columns), or pKFSS1 containing intact bosRR39K (pJS239,
bosRR39K; grey columns) were assayed for their sensitivity to t-butyl hydroperoxide.
B. Protein levels in strains complemented with bosR or bosRR39K alleles. The bosRR39K::kan
R
mutant strain was transformed with the shuttle
vector pKFSS1 alone (lane 2) or was complemented with either the intact bosRR39K allele (lane 3) or with the LP bosR allele (lane 4). Lane 1
is the protein lysate from the HP B. burgdorferi parent (with the bosRR39K allele). Top, Coomassie blue stain; bottom, Western immunoblot probed
with B. burgdorferi anti-BosR.
Dilution factor
(log
10
)
supporting
growth
Redox regulation in B. burgdorferi 1357
© 2004 Blackwell Publishing Ltd, Molecular Microbiology, 54, 13521363
allele encoded by pJAB3, cells co-harbouring a napA P/
O–lacZ fusion had 250 units of b-galactosidase activity
(Fig. 6, –IPTG, white bar). When LP bosR was induced in
these cells, b-galactosidase activity increased to ª4000
units (Fig. 6, +ITPG, grey bar) indicating that BosR
exerted a positive regulatory effect on the napA P/O–lacZ
transcriptional fusion. However, in cells harbouring napA
P/O–lacZ and IPTG-inducible bosRR39K (encoded by
pJAB3R39K), b-galactosidase activity was unchanged
independent of IPTG induction (– and +IPTG, striped and
black bars respectively) and was approximately equivalent
to activity measured in cells harbouring napA P/O–lacZ
with uninduced bosR (Fig. 6, white bar), indicating that
BosRR39K was unable to activate transcription of the
napA P/O–lacZ transcriptional fusion. Immunoblot analy-
sis demonstrated that BosRR39K was synthesized at
greater levels in cells induced with IPTG (Fig. 6, inset,
lane 2), compared with uninduced cells (Fig. 6, inset, lane
1), indicating that, unlike BosR, increases in the level of
BosRR39K did not stimulate expression from the napA P/
O–lacZ reporter (Boylan et al., 2003).
There are several possibilities that might explain why
BosRR39K does not activate the napA P/O–lacZ reporter
construct. One possible explanation is that BosRR39K
has reduced affinity for the napA promoter and thus an
impaired ability to activate transcription. To determine
whether BosRR39K could bind to the napA P/O region,
purified BosRR39K or LP BosR was incubated with the
napA P/O target sequence in the presence of 10 mM Zn
2+
and 1 mM dithiothreitol and analysed by gel electrophore-
sis. As a positive control, BosR was incubated with the
napA P/O probe and, at concentrations starting at
125 nM, an electrophoretic shift pattern was initiated iden-
tical to previous observations (Fig. 7C; Boylan et al.,
2003). In contrast, BosRR39K did not bind to the napA
P/O probe at any of the concentrations tested (up to
1000 nM; Fig. 7A). As such, the R39K mutation has dra-
matically altered the affinity of this regulatory protein to
bind the napA P/O region.
Evidence for repressor activity of BosRR39K
Inasmuch as the inability of BosRR39K to bind to the
napA P/O region (Fig. 7A) provided little information into
how the bosRR39K::kan
R
strain was hyper-resistant to
ROS (Figs 3, 4 and 5A), an additional explanation for the
increased resistance to ROS in the bosRR39K::kan
R
mutant might result from the derepression of other target
sites that result in the increased synthesis of enzymes that
neutralize these toxic compounds. Along these lines, we
had already observed an increase in SodA levels in the
bosRR39K::kan
R
background relative to its isogenic par-
ent implying that BosRR39K was functioning as a repres-
sor for the sodA locus (Fig. 3B). To test this hypothesis,
we incubated increasing amounts of BosRR39K protein
with a DNA fragment containing the putative P/O region
of sodA and found that BosRR39K bound to the DNA
fragment at concentrations starting at 125 nM (Fig. 7B) as
Fig. 6. Recombinant BosR and BosRR39K are functionally distinct.
Induction of LP wild-type bosR (+IPTG) activates transcription of lacZ
when fused to the B. burgdorferi napA promoter (napA P/O–lacZ) in
E. coli (dark grey bar) whereas induced HP bosRR39K (black bar) is
unable to despite increased synthesis of HP BosRR39K after induc-
tion (inset). The y-axis indicates the b-galactosidase (b-gal) activity
in Miller units. Error bars indicate standard deviation.
b-Gal
activity
Fig. 7. BosRR39K does not bind to the napA promoter/operator (P/
O) domain but does bind to the sodA P/O domain under conditions
where the BosR protein exhibits high affinity binding to both targets.
BosRR39K (A and B) and BosR (C and D) were incubated with
32
P-
labelled DNA containing either the napA P/O region (A and C) or the
sodA P/O domain (B and D) as indicated, separated by electrophore-
sis, and autoradiographed. Concentrations of BosR or BosRR39K
used for the napA gel shifts (A and C) were 62.5, 125, 250, 500 and
1000 nM. Concentrations for BosR and BosRR39K used for the sodA
P/O gel shifts (C and D) were the same as the napA gel shifts with
an additional 1500 nM sample. As controls, the first lane in each
panel has no added BosR or BosRR39K respectively.
1358 J. Seshu et al.
© 2004 Blackwell Publishing Ltd, Molecular Microbiology, 54, 13521363
was seen with the BosR binding to the napA P/O-contain-
ing fragment (Fig. 7C). In addition, we tested the ability of
BosR to bind to the sodA P/O domain (Fig. 7D) and found
that this region was recognized in a manner commensu-
rate with the napA fragment tested (Fig. 7C). Interestingly,
the binding properties of BosR to the sodA P/O domain
(Fig. 7D) was distinct from that seen for BosRR39K
(Fig. 7B) with BosR eliciting a more dramatic gel shift
relative to BosRR39K. These results suggest that the
R39K mutation alters the ability of BosRR39K to interact
with DNA targets thereby either reducing or eliminating its
ability to impede the migration of the sodA P/O (Fig. 7B)
or napA P/O fragment (Fig. 7A) in the gel matrix respec-
tively. In the case of the sodA P/O domain, this binding
presumably results in repression of sodA. Therefore, in
the absence of BosRR39K (i.e. the bosRR39K::kan
R
strain) sodA is derepressed resulting in increased super-
oxide dismutase levels (Fig. 3B) which are, in part,
responsible for the significantly enhanced resistance to
oxidative stressors observed (see Figs 3–5).
Induction of additional antigens independent of
oxidative stress
If BosR regulates a group of unlinked genes (i.e. consti-
tutes a regulon or stimulon), then it is likely that BosR
would concurrently regulate other additional genes whose
function may not be linked to resistance to oxidative stres-
sors. This characteristic may also be applicable to the
BosRR39K-regulated genes. To assess this possibility, we
analysed the total protein profile of the bosRR39K::kan
R
relative to its isogenic parent and determined that a faint
20 kDa protein was evident in strains lacking BosRR39K.
We then utilized anti-sera to proteins in the 20 kDa size
range in Western immunoblot analyses to ascertain the
identity of the 20 kDa species. In the subsequent analy-
ses, we determined that levels of NapA protein were sim-
ilar in the HP B. burgdorferi and the bosRR39K::kan
R
(JS167) cells (data not shown). Two additional candidates
for the 20 kDa protein were the lipoproteins DbpA and
OspC. DbpA is a well-characterized adhesin to host deco-
rin (Guo et al., 1998) whereas OspC is required for move-
ment of B. burgdorferi into the mammalian hosts in a
process that may (Pal et al., 2004) or may not (Grimm
et al., 2004) involve invasion of B. burgdorferi into the
salivary glands of Ixodes ticks. Subsequent immunoblots
with appropriate anti-sera indicated that, while OspC lev-
els are unaffected by the absence of BosRR39K (data not
shown), the levels of DbpA were significantly increased
(Fig. 8). Subsequent twofold dilutions of whole-cell lysates
resolved by SDS-PAGE and immunoblotted indicated
that DbpA was synthesized fourfold more in the
bosRR39K::kan
R
background (data not shown). This
result provides additional suggestive evidence that BosR
modulates a regulon of genes required for resistance to
oxidative stressors and perhaps other genes like dbpA
whose functions are apparently unlinked from the oxida-
tive stress response. Thus, regulation via BosR may also
be important in both host adaptation and pathogenic
mechanisms.
Discussion
The response to ROS such as superoxide anion, hydroxyl
radicals and hydrogen peroxide is important to the survival
of all living systems as these toxic molecules can react
with and damage DNA, proteins or lipids. Several prokary-
otes utilize regulatory systems that directly respond to
increased levels of ROS and/or divalent cations to combat
the toxic effects of these molecules. For example, in
E. coli OxyR, the oxidative stress regulatory protein
responds to ROS (particularly H
2
O
2
) by activating the
expression of protective enzymes such as AhpR (alkyl
hydroperoxide reductase), KatG (catalase) and GorA (glu-
tathione reductase) based on the redox status of con-
served cysteine residues (Zheng et al., 1998; Storz and
Imlay, 1999; Choi et al., 2001). The regulation of the oxyR
regulon is complex and involves both transcriptional acti-
vation and repression as well as cross-talk with the metal-
dependent repressor, Fur, and oxyS, a regulatory RNA
(Altuvia et al., 1997; Zheng et al., 1999; 2001). In Bacillus
subtilis, the cellular response to ROS is modulated by the
metalloprotein, PerR (Chen et al., 1995; Mongkolsuk and
Helmann, 2002) in a process that presumably is mediated
by the redox status of one of the conserved C–X–X–C
motifs. B. subtilis PerR regulation is complex and the ROS
sensing form of the protein contains structural Zn and
reactive Fe (Herbig and Helmann, 2001). A Zn/Mn form
Fig. 8. Decorin binding protein A is synthesized at greater levels in
the bosRR39K::kan
R
mutant. Total-cell lysates from HP B. burgdorferi
(lane 1) and the bosRR39K::kan
R
mutant (lane 2) were resolved by
SDS-PAGE and (A) stained with Coomassie blue or (B) immunoblot-
ted and probed with anti-DbpA. Markers (in kDa) are indicated on the
left.
Redox regulation in B. burgdorferi 1359
© 2004 Blackwell Publishing Ltd, Molecular Microbiology, 54, 13521363
of B. subtilis PerR was also described and both forms
interact with P/O sequences upstream of genes in the
PerR regulon (e.g. ahpCF, katG) with different affinities
and slightly different target sequences (Herbig and Hel-
mann, 2001; Mongkolsuk and Helmann, 2002). As more
is learned about these and other ROS regulatory net-
works, it is becoming clear that their regulators are multi-
functional and have global affects on gene regulation and
pathogenesis.
Recently, Boylan et al. (2003) described a Borrelia oxi-
dative stress regulator, designated BosR. The purified
BosR protein required Zn for efficient binding to a target
sequence upstream of the gene encoding the putative
oxidative stress protein, NapA. Subsequent biochemical
analysis revealed that BosR bound the napA upstream
target with higher affinity when challenged with ROS (Boy-
lan et al., 2003). This is in contrast to B. subtilis PerR
whose binding was abolished by oxidizing agents like
H
2
O
2
and t-butyl hydroperoxide (Herbig and Helmann,
2001). Transcriptional fusion data suggested that the pres-
ence of ROS increased borrelial napA expression in a
fashion that was indicative of a transcriptional activator.
Additionally, DNAase protection assays indicated that the
protein bound a 50 bp region 130 bp upstream of the
transcriptional start site, suggesting that BosR activated
transcription from a remote site in relation to the promoter
(Boylan et al., 2003). Zheng et al. (2001) had identified
an OxyR target sequence in E. coli, ª323 nucleotides
upstream of previously identified transcription start site
(P1) of ahpC. Subsequent primer extension and Northern
analysis revealed that this binding site recruited RNA poly-
merase and activated transcription from a second pro-
moter (designated P2). Currently, it is not known how
BosR activates transcription from the binding sequence
upstream of napA or whether this type of regulation
occurs for other borrelial genes, including sodA. The iden-
tification and characterization of these target sites is of
considerable interest inasmuch as the genome of B. burg-
dorferi predicts only a few genes involved in oxidative
stress detoxification (i.e. napA and sodA) with several
important omissions (Fraser et al., 1997). Specifically, B.
burgdorferi is missing genes encoding catalase, peroxi-
dases and glutathione synthesis/redox recycling, suggest-
ing that this spirochetal bacterium utilizes novel means to
neutralize ROS.
The two alleles of bosR identified in this study (i.e. bosR
and bosRR39K) provide some new insights as to how B.
burgdorferi BosR may function as a transcriptional regu-
lator. Our studies focused on an HP B. burgdorferi non-
infectious isolate which had increased sensitivity to ROS
when compared with LP infectious and other HP isolates
(Fig. 1). Insertional disruption of bosRR39K allele in the
HP strain (Fig. 2) resulted in increased resistance to ROS,
suggesting that BosRR39K functioned as a transcriptional
repressor (Figs 3–5). Genetic complementation with the
isogenic allele (bosRR39K) restored sensitivity to t-butyl
hydroperoxide in the mutant strain indicating that the phe-
notype of the mutant was attributed specifically to the loss
of BosRR39K (Fig. 5A). However, when the LP wild-type
bosR allele was used to complement the disrupted HP
bosRR39K::kan
R
mutant, an intermediate sensitivity to
ROS was observed. Additionally, activity gels and immu-
noblots indicated that sodA and dbpA were expressed at
higher levels in the mutant strain (Figs 3B and 8 respec-
tively) whereas napA expression remained unchanged
(data not shown). Taken together, these data suggest that
the bosR and bosRR39K alleles are functionally distinct.
In addition, although additional secondary mutations con-
tribute in part to the hypersensitivity to ROS observed for
our HP B. burgdorferi strain and were instrumental in the
isolation of the bosRR39K genetic knockout inasmuch as
bosR knockouts cannot be isolated under these experi-
mental conditions, our data suggest that the effect of the
R39K mutation on bosR is in good measure responsible
for the phenotype observed particularly in regard to
increased sensitivity to oxidative stressors. Reporter con-
structs to decipher the differences in regulatory activity for
BosR and BosRR39K are currently being designed and
evaluated.
The difference in gel shift of the sodA fragment seen for
BosR and BosRR39K implies that the R39K mutation
affects oligomerization such that BosRR39K can still bind
to the sodA P/O but does not impede migration in the gel
matrix to the extent observed for BosR (compare Fig. 7B
and D). Oligomerization of BosR would not be without
precedent as other Fur family regulators form oligomeric
structures that are important for their regulatory activity
(Escolar et al., 1999; Lavrrar et al., 2002; Friedman
and O’Brian, 2003; Pohl et al., 2003). The inability of
BosRR39K to oligomerize may also explain the lack of
binding to the napA P/O (Fig. 7A) if one assumes that a
higher ordered structure is required for the BosR-depen-
dent gel shift observed (Fig. 7C). Thus, although the direct
effect of the mutation is not known as yet, the phenotypes
observed imply that residue 39 is key for BosR function.
Recently, Seshu et al. (2004) reported that DbpA levels
increased in response to decreased oxygen levels imply-
ing that the redox status of B. burgdorferi can modulate
gene expression. There are several possible explanations
for this response. First, BosR may modulate its regulatory
activity presumably via the oxidation of one of its two C–
X–X–C motif(s) using a mechanism not unlike OxyR
(Zheng et al., 1998; Storz and Imlay, 1999; Choi et al.,
2001). In this model, the redox status of BosR dictates the
regulatory activity of this protein. The studies presented
by Boylan et al. (2003) and herein indicate that reduced
BosR is able to bind either the napA or sodA P/O regions
whereas BosRR39K only binds to the sodA P/O fragment
1360 J. Seshu et al.
© 2004 Blackwell Publishing Ltd, Molecular Microbiology, 54, 13521363
(Fig. 7). In contrast, oxidized BosR bound more avidly
than reduced BosR and the redox-dependent activation
via BosR of a napA P/O–lacZ transcriptional fusion in
E. coli (Boylan et al., 2003) suggests that, for some target
genes, oxidation of BosR may activate transcription.
Whether subsequent oxidation affects the ability of
BosRR39K to bind DNA or activate transcription at target
genes remains to be determined. Second, recent studies
indicated that dbpA is regulated by RpoS via RpoN-
dependent activation (Hübner et al., 2001) in conjunction
with the response regulator Rrp2 (Yang et al., 2003)
whose cognate sensor kinase contains a PAS domain.
Inasmuch as PAS motifs have been shown to sense the
redox status of cells (along with other signals), it is pos-
sible that the sensor kinase transduces a redox-based
signal resulting in increased expression of dbpA and
ospC. Whether the BosR-dependent regulation and the
two-component systems are subject to molecular cross-
talk remains to be determined.
It is well established that B. burgdorferi alters gene
expression in response to the different signals that reflect
changes in their ‘environment’ as the bacterium moves
from ticks to mammalian hosts. In particular, it has been
shown that growth temperature, pH and host adaptation
dramatically affect the expression of ospC, dbpA and
other genes (Schwan et al., 1995; Akins et al., 1998; Car-
roll et al., 1999; Yang et al., 2000; Brooks et al., 2003).
More recently, the redox status of B. burgdorferi was also
shown to modulate some of these same genes (Seshu
et al., 2004). Based on the aforementioned studies, it is
becoming clear that regulation of oxidative stress genes
by BosR represents an additional adaptive response that
might be important for B. burgdorferi survival during the
initial stages of infection. For example, after B. burgdorferi
enters the infection site, it would subsequently be chal-
lenged by innate host defences including challenge by
reactive oxygen and nitrogen species released by neutro-
phils. As such, the regulation mediated by BosR may be
important in initiating a molecular cascade that protects
B. burgdorferi from oxidative damage at these sites. The
exact role of BosR and the genes it regulates in the
context of Lyme disease pathogenesis remains to be
determined experimentally.
Experimental procedures
Bacterial strains and growth conditions
Clonal isolates of HP non-infectious B. burgdorferi sensu
stricto strain B31 and LP infectious B. burgdorferi strain B31
derivative MSK5 (Labandeira-Rey and Skare, 2001) were
grown microaerophilically in modified Barbour-Stoenner-Kelly
(BSK-II) liquid or solid medium as previously described (Sam-
uels, 1995). E. coli strains Top10 and Top10F¢ (Invitrogen)
were grown in LB medium supplemented with ampicillin
(100 mg ml
-1
), kanamycin (50 mg ml
-1
) or spectinomycin
(100 mg ml
-1
). All chemicals were purchased from Sigma
Chemicals and DNA restriction or modifying enyzmes were
purchased from either New England Biolabs or Promega
unless stated otherwise.
PCR
PCR, the preparation of templates and subsequent cloning
steps were performed as described previously (Labandeira-
Rey and Skare, 2001).
Mutagenesis of bosR
To generate a mutagenesis system to insert a selectable
marker (P
flgB
-kan
R
) into B. burgdorferi genes, we have gener-
ated a plasmid construct to insertionally disrupt genes by in
vitro transposition (see Fig. 9 for a schematic of the plasmid
constructs). The resulting constructs were then used to inac-
tivate genes in B. burgdorferi by allelic exchange. To generate
the transposable P
flgB
-kan
R
, we did the following: the kan
R
determinant from Tn903 was amplified from pENT3 (kindly
provided by Brian Akerley, University of Massachusetts Med-
ical School, Worcester) by PCR using primers kan903/Mlu/F
(A
CGCGTTA ATACAAGGGGTGTTATGA) and kan903/Mlu/R
(A
CGCGTCGAGGATCCCCGCCACGGTT) generating a pro-
moterless kan
R
with MluI restriction sites (underlined). The
PCR product was cloned into pCR2.1 (Invitrogen) and
digested with MluI, and the resulting fragment was cloned
into pMT102 generating pMT103. The B. burgdorferi flgB
promoter was amplified by PCR using primers flgB/EcoRV
(GA
TATCGGAAGATTTCCTATTAAGG) and flgB/SalI (GTCG
ACCATTTAAAATTGCTTTTAAC) generating a product with
unique EcoRV and SalI restriction sites (underlined). The
resulting PCR product was digested with EcoRV and SalI and
cloned into pMT103 upstream of the promoterless kan
R
gene,
generating the plasmid pBF108. A 1480 bp EcoRI/BamHI
fragment containing P
flgB
-kan
R
was cloned into the EZ::TN
pMOD <MCS> vector (Epicentre Technologies) generating
plasmid pJS100. A 1570 bp PvuII fragment was excised from
pJS100 and ligated into pGPS3 (New England Biolabs) gen-
erating pJS104 that contains P
flgB
-kan
R
flanked by both Tn5
and Tn7 transposon ends.
To generate the target for in vitro transposition, a 4978 bp
fragment from LP B. burgdorferi strain B31 isolate containing
bosR was amplified by PCR with primers 645 (ATGTT
AAAGGGTTTTGAACAA) and 648 (ATTACTTCTATGCATC
AAAAAATAC) and cloned into pGEM-T-Easy (Promega),
generating pJS162. In vitro mutagenesis was carried out
using the GPS-Mutagenesis System (New England Biolabs)
according to the manufacturer’s instructions with pJS162 as
the target DNA and pJS104 as the donor. Transposed con-
structs were electroporated into E. coli strain Top10 and
selected for amp
R
(100 mg ml
-1
) and kan
R
(50 mg ml
-1
). Plas-
mids from the transformants were dideoxy sequenced to
determine the exact site of transposition.
Plasmids harbouring bosR::kan
R
were transformed into
electrocompetant HP B. burgdorferi strain B31 (containing
the bosRR39K allele) as described by Samuels (1995)
except that the final wash step and resuspension was per-
Redox regulation in B. burgdorferi 1361
© 2004 Blackwell Publishing Ltd, Molecular Microbiology, 54, 13521363
formed in 8 mM Hepes pH 7.4, 272 mM sucrose. After elec-
troporation and an overnight incubation, samples were
placed in BSK-II agarose and plated as overlays onto BSK-
II plates both containing 200 mg ml
-1
kanamycin. Plates were
incubated at 32C in 1% CO
2
for 10–13 days until individual
subsurface colonies were visible. Colonies were aseptically
picked and inoculated into BSK-II liquid media containing
200 mg ml
-1
kanamycin. Primers bosR/BamHI (GGATCCAT G
AACGACAACATAATAAG) and bosR/PstI (CTGCA
GT
AAAGTGATTTCCTTGTTCT) were used to PCR amplify the
B. burgdorferi bosRR39K allele to confirm the presence of
the P
flgB
-kan
R
cassette in our HP isolate. Several clones con-
taining bosRR39K::kan
R
were obtained and one, designated
JS167 was used for the studies described below.
Genetic complementation
To complement the HP bosRR39K::kan
R
mutant, bosRR39K
and bosR were amplified from HP strain B31 and LP strain
MSK5, respectively, using primers bosR/com/Bam (GGA
TCC
TGCTCCAAATCCATGAATA) and bosR/com/Pst (CTGC
AGTTTAAATGTTGAAAAACATA). Each PCR product was
cloned into pKFSS1 (Frank et al., 2003; generously provided
by D. Scott Samuels, University of Montana), which confers
resistance to streptomycin in B. burgdorferi, generating
pJS232 (which contains the bosR allele) and pJS239 (which
contains the bosRR39K allele). After sequencing to confirm
the presence of the appropriate bosR allele, each construct
was electroporated into B. burgdorferi strain JS167 as
described above and the transformants were plated in
agarose overlays on BSK-II plates both containing 50 mg
ml
-1
streptomycin. As a control, pKFSS1 was also electropo-
rated into B. burgdorferi strain JS167. Isolates were analysed
by PCR with primers bosR/BamHI and bosR/PstI to verify
that either bosR or bosRR39K were coresident in the
bosRR39K::kan
R.
background JS167.
Sensitivity of B. burgdorferi constructs to ROS
Cultures of LP B. burgdorferi strain B31 derivative MSK5
(with the bosR allele), our HP B. burgdorferi strain B31
(with the bosRR39K allele), JS167 (bosRR39K::kan
R
),
JS167/pKFSS1 (bosRR39K::kan
R
/vector control), JS167/
pJS232 (bosRR39K::kan
R
/bosR) and JS167/pJS239
(bosRR39K::kan
R
/bosRR39K), were grown microaerophili-
cally to a cell density of 5 ¥ 10
7
cells per ml in BSK-II media
and harvested by centrifugation (8000 g, 10 min, room tem-
perature). In all cases, the strains tested were grown inde-
pendently and assayed in triplicate. Cells were suspended in
PBS/0.4% glucose/2% bovine serum albumin (PBS-BSA-G)
Fig. 9. Schematic of donor plasmid constructs for in vitro transposition. The construction of the custom transposon utilized in this study is described
in Experimental procedures. The final construct, pJS104, contains both Tn5 and Tn7 ends flanking the P
flgB
-kan
R
cassette. The total length of the
transposable unit between the Tn7 ends is 2026 bp. TnL and TnR represent the mos1 transposon ends; blocks labelled Tn5 and Tn7 (Tn7L and
Tn7R) represent the appropriate transposon ends respectively. Amp
R
, ampicillin resistance gene; kan
R
, kanamycin resistance gene; ColEI ori,
ColEI origin of replication; P
flgB
, promoter sequence from flgB gene of B. burgdorferi strain B31.
1362 J. Seshu et al.
© 2004 Blackwell Publishing Ltd, Molecular Microbiology, 54, 13521363
or PBS-BSA-G that contained either H
2
O
2
or t-butyl hydrop-
eroxide at concentrations ranging from 0 to 4 mM, and incu-
bated for 1 h at 32C. After incubation, the samples were
washed twice, serially diluted fivefold in BSK-II media using
a 96-well plate, and the incubation continued at 32C for
3 weeks. Growth was scored over a 3 week period by dark
field microscopy. For methyl viologen assays, cells were sus-
pended in BSK-II media. All data shown were derived from
three independent assays. All data were converted to loga-
rithmic values before the calculation of averages and statis-
tical analysis. The Student’s t-test was used to compare
differential sensitivities of B. burgdorferi isolates to various
ROS. Significant differences were accepted when the P-
value was <0.05.
Mobility shift DNA-binding assays
Recombinant BosR and BosRR39K were purified and gel
shift assays were performed as previously described (Boylan
et al., 2003). The targets for BosR and BosRR39K binding
correspond to the following putative P/O regions of the fol-
lowing borrelial genes: sodA, chromosomal co-ordinates
153 953–154 352 (within the 3¢ domain of secA); napA, as
previously described (Boylan et al., 2003). Oligonucleotide
primers (18–21 nucleotides in length) were designed to the
domains listed above and the corresponding putative P/O
domains were PCR amplified and end labelled with [g-
32
P]-
ATP as previously described (Boylan et al., 2003).
Electrophoresis and immunoblotting
Southern blots were performed as described previously
(Skare et al., 1999). For protein separations, samples were
resolved on SDS-12.5% polyacrylamide gels. Immunoblot-
ting was performed by transferring SDS-PAGE samples to
either nitrocellulose (0.2 mm Trans-Blot
®
, Bio-Rad) or PVDF
membranes (0.45 mm, Millipore) blocked with 5% non-fat dry
milk in PBS (12 h, 4C) and probed with BosR anti-serum
(generated against purified BosR by Cocalico; Boylan et al.,
2003) or with anti-DbpA anti-serum as previously described
(Skare et al., 1999). The blots were incubated with horserad-
ish peroxidase-coupled anti-rabbit Ig secondary antibodies
and immunoreactive bands were visualized by chemilumines-
cence (ECL Plus
®
, Amersham Pharmacia).
Enzyme assays
b-Galactosidase activity from E. coli harbouring transcrip-
tional fusions was measured as described (Boylan et al.,
2003). E. coli cells harbouring napA P/O–lacZ fusion cores-
ident with pJAB3 (contains IPTG-inducible bosR; Boylan
et al., 2003) or pJAB3R39K (identical construct as pJAB3
except that the bosRR39K allele replaces bosR) were grown
in M9 minimal media or M9 + IPTG at 37C to mid-log phase
(OD
600
ª0.4–0.5). Cells were harvested by centrifugation
(5000 g, 5 min, 4C) and assayed for enzyme activity as
previously described (Boylan et al., 2003). For SodA analy-
ses, B. burgdorferi cells were grown under microaerophilic
conditions and the cells were grown to mid-log phase and
lysed by sonication. The presence of superoxide dismutase
activity from the resulting B. burgdorferi lysates was deter-
mined as previously described (Beauchamp and Fridovich,
1971).
Acknowledgements
We thank Magnus Höök, Texas A&M University HSC, Institute
of Biosciences and Technology, Houston, for providing anti-
serum to DbpA. We are also grateful to D. Scott Samuels,
University of Montana, for providing pKFSS1, and Brian Aker-
ley, University of Massachusetts Medical School, Worcester
for providing us with plasmid pENT3. We also thank Maria
Labandeira-Rey for valuable and helpful comments and
Dolores Esteve-Gassent for critical evaluation of this manu-
script. This work was supported by Public Health Service
Grant AI42345 from the National Institute of Allergy and
Infectious Diseases (to J.T.S.).
References
Akins, D.R., Bourell, K.W., Caimano, M.J., Norgard, M.V.,
and Radolf, J.D. (1998) A new animal model for studying
Lyme disease spirochetes in a mammalian host-adapted
state. J Clin Invest 101: 2240–2250.
Altuvia, S., Weinstein-Fischer, D., Zhang, A., Postow, L., and
Storz, G. (1997) A small, stable RNA induced by oxidative
stress: role as a pleiotropic regulator and antimutator. Cell
90: 43–53.
Beauchamp, C., and Fridovich, I. (1971) Superoxide dismu-
tase: improved assays and an assay applicable to acryla-
mide gels. Anal Biochem 44: 276–287.
Bono, J.L., Elias, A.F., Kupko, J.J., 3rd, Stevenson, B., Tilly,
K., and Rosa, P. (2000) Efficient targeted mutagenesis in
Borrelia burgdorferi. J Bacteriol 182: 2445–2452.
Boylan, J.A., Posey, J.E., and Gherardini, F.C. (2003) Borre-
lia oxidative stress response regulator, BosR: a distinctive
Zn-dependent transcriptional activator. Proc Natl Acad Sci
USA 100: 11684–11689.
Brooks, C.S., Hefty, P.S., Jolliff, S.E., and Akins, D.R. (2003)
Global analysis of Borrelia burgdorferi genes regulated by
mammalian host-specific signals. Infect Immun 71: 3371–
3383.
Carroll, J.A., Garon, C.F., and Schwan, T.G. (1999) Effects
of environmental pH on membrane proteins in Borrelia
burgdorferi. Infect Immun 67: 3181–3187.
Chen, L., Keramati, L., and Helmann, J.D. (1995) Coordinate
regulation of Bacillus subtilis peroxide stress genes by
hydrogen peroxide and metal ions. Proc Natl Acad Sci USA
92: 8190–8194.
Choi, H., Kim, S., Mukhopadhyay, P., Cho, S., Woo, J., Storz,
G., and Ryu, S. (2001) Structural basis of the redox switch
in the OxyR transcription factor. Cell 105: 103–113.
Das, S., Barthold, S.W., Giles, S.S., Montgomery, R.R., Tel-
ford, S.R.R., and Fikrig, E. (1997) Temporal pattern of
Borrelia burgdorferi p21 expression in ticks and the mam-
malian host. J Clin Invest 99: 987–995.
Escolar, L., Perez-Martin, J., and de Lorenzo, V. (1999)
Opening the iron box: transcriptional metalloregulation by
the Fur protein. J Bacteriol 181: 6223–6229.
Frank, K.L., Bundle, S.F., Kresge, M.E., Eggers, C.H., and
Samuels, D.S. (2003) aadA confers streptomycin
Redox regulation in B. burgdorferi 1363
© 2004 Blackwell Publishing Ltd, Molecular Microbiology, 54, 13521363
resistance in Borrelia burgdorferi. J Bacteriol 185: 6723–
6727.
Fraser, C.M., Casjens, S., Huang, W.M., Sutton, G.G., Clay-
ton, R., Lathigra, R., et al. (1997) Genomic sequence of a
Lyme disease spirochaete, Borrelia burgdorferi. Nature
390: 580–586.
Friedman, Y.E., and O’Brian, M.R. (2003) A novel DNA-
binding site for the ferric uptake regulator (Fur) protein from
Bradyrhizobium japonicum. J Biol Chem 278: 38395–
38401.
Fuangthong, M., Herbig, A.F., Bsat, N., and Helmann, J.D.
(2002) Regulation of the Bacillus subtilis fur and perR
genes by PerR: not all members of the PerR regulon are
peroxide inducible. J Bacteriol 184: 3276–3286.
Grimm, D., Tilly, K., Byram, R., Stewart, P.E., Krum, J.G.,
Bueschel, D.M., et al. (2004) Outer-surface protein C of
the Lyme disease spirochete: a protein induced in ticks for
infection of mammals. Proc Natl Acad Sci USA 101: 3142–
3147.
Guo, B.P., Brown, E.L., Dorward, D.W., Rosenberg, L.C., and
Höök, M. (1998) Decorin-binding adhesins from Borrelia
burgdorferi. Mol Microbiol 30: 711–723.
Herbig, A.F., and Helmann, J.D. (2001) Roles of metal ions
and hydrogen peroxide in modulating the interaction of the
Bacillus subtilis PerR peroxide regulon repressor with oper-
ator DNA. Mol Microbiol 41: 849–859.
Hübner, A., Yang, X., Nolen, D.M., Popova, T.G., Cabello,
F.C., and Norgard, M.V. (2001) Expression of Borrelia
burgdorferi OspC and DbpA is controlled by a RpoN-RpoS
regulatory pathway. Proc Natl Acad Sci USA 98: 12724–
12729.
Labandeira-Rey, M., and Skare, J.T. (2001) Decreased infec-
tivity in Borrelia burgdorferi strain B31 is associated with
loss of linear plasmid 25 or 28-1. Infect Immun 69: 446–
455.
Labandeira-Rey, M., Seshu, J., and Skare, J.T. (2003) The
absence of linear plasmid 25 or 28-1 of Borrelia burgdorferi
dramatically alters the kinetics of experimental infection via
distinct mechanisms. Infect Immun 71: 4608–4613.
Lavrrar, J.L., Christoffersen, C.A., and McIntosh, M.A. (2002)
Fur–DNA interactions at the bidirectional fepDGC-entS
promoter region in Escherichia coli. J Mol Biol 322: 983–
995.
McDowell, J.V., Sung, S.Y., Labandeira-Rey, M., Skare, J.T.,
and Marconi, R.T. (2001) Analysis of mechanisms associ-
ated with loss of infectivity of clonal populations of Borrelia
burgdorferi B31MI. Infect Immun 69: 3670–3677.
Mongkolsuk, S., and Helmann, J.D. (2002) Regulation of
inducible peroxide stress responses. Mol Microbiol 45: 9–
15.
Ohnishi, J., Piesman, J., and de Silva, A.M. (2001) Antigenic
and genetic heterogeneity of Borrelia burgdorferi popula-
tions transmitted by ticks. Proc Natl Acad Sci USA 98:
670–675.
Ojaimi, C., Brooks, C., Casjens, S., Rosa, P., Elias, A.,
Barbour, A., et al. (2003) Profiling of temperature-
induced changes in Borrelia burgdorferi gene expression
by using whole genome arrays. Infect Immun 71: 1689–
1705.
Pal, U., Yang, X., Chen, M., Bockenstedt, L.K., Anderson,
J.F., Flavell, R.A., et al. (2004) OspC facilitates Borrelia
burgdorferi invasion of Ixodes scapularis salivary glands. J
Clin Invest 113: 220–230.
Pohl, E., Haller, J.C., Mijovilovich, A., Meyer-Klaucke, W.,
Garman, E., and Vasil, M.L. (2003) Architecture of a pro-
tein central to iron homeostasis: crystal structure and spec-
troscopic analysis of the ferric uptake regulator. Mol
Microbiol 47: 903–915.
Posey, J.E., and Gherardini, F.C. (2000) Lack of a role for
iron in the Lyme disease pathogen. Science 288: 1651–
1653.
Purser, J.E., and Norris, S.J. (2000) Correlation between
plasmid content and infectivity in Borrelia burgdorferi. Proc
Natl Acad Sci USA 97: 13865–13870.
Purser, J.E., Lawrenz, M.B., Caimano, M.J., Howell, J.K.,
Radolf, J.D., and Norris, S.J. (2003) A plasmid-encoded
nicotinamidase (PncA) is essential for infectivity of Borrelia
burgdorferi in a mammalian host. Mol Microbiol 48: 753–
764.
Revel, A.T., Talaat, A.M., and Norgard, M.V. (2002) DNA
microarray analysis of differential gene expression in Bor-
relia burgdorferi, the Lyme disease spirochete. Proc Natl
Acad Sci USA 99: 1562–1567.
Samuels, D.S. (1995) Electrotransformation of the spirochete
Borrelia burgdorferi. Methods Mol Biol 47: 253–259.
Schwan, T.G., Piesman, J., Golde, W.T., Dolan, M.C., and
Rosa, P.A. (1995) Induction of an outer surface protein on
Borrelia burgdorferi during tick feeding. Proc Natl Acad Sci
USA 92: 2909–2913.
Seshu, J., Boylan, J.A., Gherardini, F.C., and Skare, J.T.
(2004) Dissolved oxygen levels alter gene expression and
antigen profiles in Borrelia burgdorferi. Infect Immun 72:
1580–1586.
Skare, J.T., Foley, D.M., Hernandez, S.R., Moore, D.C.,
Blanco, D.R., Miller, J.N., and Lovett, M.A. (1999) Cloning
and molecular characterization of plasmid-encoded anti-
gens of Borrelia burgdorferi. Infect Immun 67: 4407–4417.
Storz, G., and Imlay, J.A. (1999) Oxidative stress. Curr Opin
Microbiol 2: 188–194.
Yang, X., Goldberg, M.S., Popova, T.G., Schoeler, G.B.,
Wikel, S.K., Hagman, K.E., and Norgard, M.V. (2000) Inter-
dependence of environmental factors influencing recipro-
cal patterns of gene expression in virulent Borrelia
burgdorferi. Mol Microbiol 37: 1470–1479.
Yang, X.F., Alani, S.M., and Norgard, M.V. (2003) The
response regulator Rrp2 is essential for the expression of
major membrane lipoproteins in Borrelia burgdorferi. Proc
Natl Acad Sci USA 100: 11001–11006.
Zheng, M., Aslund, F., and Storz, G. (1998) Activation of the
OxyR transcription factor by reversible disulfide bond for-
mation. Science 279: 1718–1721.
Zheng, M., Doan, B., Schneider, T.D., and Storz, G. (1999)
OxyR and SoxRS regulation of fur. J Bacteriol 181: 4639–
4643.
Zheng, M., Wang, X., Doan, B., Lewis, K.A., Schneider, T.D.,
and Storz, G. (2001) Computation-directed identification of
OxyR DNA binding sites in Escherichia coli. J Bacteriol
183: 4571–4579.
    • "Although both rpoN and rpoS showed changes in gene expression in response to lower osmolarity, the sigma factor, rpoD, did not change in response to osmolarity (Fig 4). Regulation of rpoS and RpoS is transcriptional, translational and post-translational [7, 8,[38][39][40][41] . Expression analysis of the Borrelia oxidative stress regulator (BosR), which is thought to directly regulate rpoS, indicated that there was no change in transcription or translation of bosR in response to changes in osmolarity (Figs 3 and 4). "
    [Show abstract] [Hide abstract] ABSTRACT: Lyme disease, caused by Borrelia burgdorferi, is a vector-borne illness that requires the bacteria to adapt to distinctly different environments in its tick vector and various mammalian hosts. Effective colonization (acquisition phase) of a tick requires the bacteria to adapt to tick midgut physiology. Successful transmission (transmission phase) to a mammal requires the bacteria to sense and respond to the midgut environmental cues and up-regulate key virulence factors before transmission to a new host. Data presented here suggest that one environmental signal that appears to affect both phases of the infective cycle is osmolarity. While constant in the blood, interstitial fluid and tissue of a mammalian host (300 mOsm), osmolarity fluctuates in the midgut of feeding Ixodes scapularis. Measured osmolarity of the blood meal isolated from the midgut of a feeding tick fluctuates from an initial osmolarity of 600 mOsm to blood-like osmolarity of 300 mOsm. After feeding, the midgut osmolarity rebounded to 600 mOsm. Remarkably, these changes affect the two independent regulatory networks that promote acquisition (Hk1-Rrp1) and transmission (Rrp2-RpoN-RpoS) of B. burgdorferi. Increased osmolarity affected morphology and motility of wild-type strains, and lysed Hk1 and Rrp1 mutant strains. At low osmolarity, Borrelia cells express increased levels of RpoN-RpoS-dependent virulence factors (OspC, DbpA) required for the mammalian infection. Our results strongly suggest that osmolarity is an important part of the recognized signals that allow the bacteria to adjust gene expression during the acquisition and transmission phases of the infective cycle of B. burgdorferi.
    Full-text · Article · Aug 2016
    • "The Hk2-Rrp2 TCS activates the expression of the stationary phase sigma factor RpoS synergistically with RpoN (Burtnick et al., 2007; Ouyang et al., 2008; Blevins et al., 2009), which, in turn, chiefly regulates plasmid-borne genes (Yang et al., 2003a,b; Caimano et al., 2007) and induces the expression of genes, such as ospC (Hübner et al., 2001), which are known to be important for mammalian infection (Caimano et al., 2004; Fisher et al., 2005; Caimano et al., 2007; Boardman et al., 2008; Ouyang et al., 2008; Dunham-Ems et al., 2012; Ouyang et al., 2012) as well as genes involved in chitobiose utilization, which has been shown to be important for colonization of the tick (Sze et al., 2013). The Hk1-Rrp1 TCS converges with the Hk2-Rrp2 TCS through the regulator, BosR—a Fur/Per-like transcription factor that has been demonstrated to be essential for expression of rpoS (Boylan et al., 2003; Katona et al., 2004; Seshu et al., 2004; Hyde et al., 2009; Ouyang et al., 2009, 2011; Hyde et al., 2010)—which primarily regulates core chromosome-encoded genes (Rogers et al., 2009; He et al., 2011, 2014) and is required for tick colonization (Caimano et al., 2011; He et al., 2011; Kostick et al., 2011). Interestingly, Rrp1, the response regulator, lacks a DNA-binding domain, but instead contains a GGDEF domain, which has been associated with diguanylate cyclase activity (cyclic di-GMP synthase) in B. burgdorferi (Ryjenkov et al., 2005). "
    [Show abstract] [Hide abstract] ABSTRACT: In nature, the Lyme disease spirochete Borrelia burgdorferi cycles between the unrelated environments of the Ixodes tick vector and mammalian host. In order to survive transmission between hosts, B. burgdorferi must be able to not only detect changes in its environment, but also rapidly and appropriately respond to these changes. One manner in which this obligate parasite regulates and adapts to its changing environment is through cyclic-di-GMP (c-di-GMP) signaling. c-di-GMP has been shown to be instrumental in orchestrating the adaptation of B. burgdorferi to the tick environment. B. burgdorferi possesses only one set of c-di-GMP-metabolizing genes (one diguanylate cyclase and two distinct phosphodiesterases) and one c-di-GMP-binding PilZ-domain protein designated as PlzA. While studies in the realm of c-di-GMP signaling in B. burgdorferi have exploded in the last few years, there are still many more questions than answers. Elucidation of the importance of c-di-GMP signaling to B. burgdorferi may lead to the identification of mechanisms that are critical for the survival of B. burgdorferi in the tick phase of the enzootic cycle as well as potentially delineate a role (if any) c-di-GMP may play in the transmission and virulence of B. burgdorferi during the enzootic cycle, thereby enabling the development of effective drugs for the prevention and/or treatment of Lyme disease.
    Full-text · Article · May 2014
    • "Cell pellets were washed twice in BSK-II medium containing 1,000 U ml 21 of catalase and resuspended in 1 ml of BSK-II medium. Quantification of viability was determined by limiting dilution in 96-well plates (Fisher Scientific, Pittsburgh, PA) following incubation at 37uC/5% CO 2 for 14 days [51]. Controls were incubated in BSK-II and BSK-II without pyruvate and treated with catalase for comparison with H 2 O 2 treated samples. "
    Full-text · Dataset · Jan 2014 · Frontiers in Cellular and Infection Microbiology
Show more