|
|
|
|||||||
|
|||||||||
(Received for publication, June 8, 1995) From the
Bacterial iron transport is critical for growth of pathogens in
the host environment, where iron is limited as a form of nonspecific
immunity. For Gram-negative bacteria such as Haemophilus
influenzae, iron first must be transported across the outer
membrane and into the periplasmic space, then from the periplasm to the
cytosol. H. influenzae express a periplasmic iron-binding
protein encoded by the hitA gene. This gene is organized as
the first of a three-gene operon purported to encode a classic high
affinity iron acquisition system that includes hitA, a
cytoplasmic permease (hitB), and a nucleotide binding protein (hitC). In this study we describe the cloning, overexpression,
and purification of the H. influenzae hitA gene product. The
function of this protein is unambiguously assigned by demonstrating its
ability to compete for iron bound to the chemical iron chelator
2,2`-dipyridyl, both in vitro and within the periplasmic space
of a siderophore-deficient strain of Escherichia coli.
Finally, the importance of a functional hitABC operon for iron
acquisition is demonstrated by complementation of this
siderophore-deficient E. coli to growth on
dipyridyl-containing medium. These studies represent a detailed
genetic, biochemical, and physiologic description of an active
transport system that has evolved to efficiently transport iron and
consequently is widely distributed among Gram-negative pathogenic
bacteria.
High affinity acquisition of iron from the host environment is a
necessary determinant of virulence for pathogenic
bacteria(1, 2, 3, 4, 5, 6, 7, 8, 9) .
This acquisition is vital for survival in the human host, where levels
of extracellular iron are tightly controlled by the Transferrins
(transferrin and lactoferrin), a family of iron-binding proteins that
function in the extracellular chelation and transport of host
iron(9) . By binding iron with high affinity, Transferrins
ensure that all extracellular iron is both efficiently sequestered from
pathogenic invaders and mobilized for transport to host tissues.
Microorganisms growing in the human host must therefore possess
mechanisms for obtaining Transferrin-sequestered iron. For a number of
pathogenic members of the Pasteurellaceae (H.
influenzae) and Neisseriaceae (Neisseria meningitidis and Neisseria gonorrhoeae), iron acquisition is initiated
by cell-surface receptors specific for the
Transferrins(10, 11, 12, 13, 14, 15, 16) .
Iron is removed from these proteins and transported across the outer
membrane, presumably by an energy-dependent TonB-mediated process (17, 18, 19) involving gated-pore properties
of the outer membrane receptor (18, 20) . The result
is deposition of free iron within the periplasm, where it is separated
from the cytosol, its eventual destination, by the cytoplasmic
membrane(21) . Transport of free iron from the periplasmic
space into the cytoplasm is proposed to occur by a classic active
transport process involving a periplasmic binding protein, a specific
cytoplasmic permease, and an energy-supplying nucleotide-binding
protein(22) . Much of what is known about the biochemistry of
active transport systems has been revealed through the study of model
active transport systems for amino acids and sugars in Escherichia
coli(22, 23) . Similar systems for
siderophore-mediated iron transport have been described for E. coli and related organisms at the genetic level(23) ; however,
relatively little is known regarding the basic biochemistry of these
iron transport processes. A genetic locus critical to the transport
of iron in H. influenzae has recently been described by Hansen
and colleagues (24) . This locus was identified through
complementation of a H. influenzae isolate unable to grow on
medium containing protoporphyrin IX and free iron. An 11.5-kb ( At the protein level,
Harkness and colleagues (27) originally observed a
quantitatively major, iron-regulated periplasmic protein, subsequently
genetically defined as hitA(24) . The predicted open
reading frame of hitA is nearly 80% homologous with the ferric
iron-binding protein (Fbp) expressed by pathogenic Neisseria(17, 29, 30, 31, 32, 33) .
Similarly, the open reading frame of sfuA predicts a protein
sequence sharing substantial homology ( Studies on the HFbp, NFbp, and
SFbp homologues and their respective operons predict that they should
function similarly. This report describes the ability of purified HFbp
to efficiently compete for dipyridyl-bound iron in vitro. Like
the sfu operon, hitABC can complement the
siderophore-deficient E. coli strain H-1443 to growth on
dipyridyl-containing medium. We further demonstrate that labeled iron
from this medium is initially bound to periplasmic HFbp and can only be
transported into the cell if a functional permease and
nucleotide-binding protein are present. These studies explain why the
siderophore-deficient E. coli expressing the hit operon can be complemented to growth on dipyridyl-containing
media. Furthermore, they represent the first comprehensive biochemical
analysis of a periplasmic iron transport system.
The plasmid pJDS150 was a
generous gift of Dr. Eric Hansen, University of Texas Southwestern.
Plasmid pBR322 was purchased from Promega (Madison, WI).
Oligonucleotides were prepared using an Applied Biosystems
International model 391 DNA Synthesizer (Foster City, CA) and were
deprotected and purified as per the manufacturer's instructions. Taq polymerase used was purchased from either Life
Technologies, Inc. or Boehringer Mannheim. Random hexamers used in
generating labeled PCR probes, T4 DNA ligase, and the restriction
enzymes EcoRI, EcoRV, BamHI, and SmaI were purchased from Boehringer Mannheim. Nutrient broth,
trypticase soy broth, components for Luria-Bertani Broth (LB), NZY
agar, Difco agar, and other media components were purchased from Difco
(Detroit, MI). Cetyltrimethylammonium bromide, 2,2`-dipyridyl,
CM-Sepharose, DEAE-Sepharose, MgS0
Figure 1:
Plasmid map of the HFbp-expressing
plasmid pBSJ1. As described in the text, an
The fragment containing the minimal hitABC from H. influenzae was prepared by PCR amplification as
described in Fig. 2. PCR reactions were performed in 100 µl
volumes using standard conditions previously described (33) and
10 units of Taq polymerase and 10 units of Taq extender. Amplification was achieved by 27 cycles of denaturation
(95 °C for 1.5 min), annealing (60 °C for 2 min), and extension
(72 °C for 3 min). At cycle 17, the reactions were replenished with
an additional 5 units of both Taq polymerase and Taq extender. Specifically, primers were designed to the extreme ends
of the hitABC sequence (24) that included
Figure 2:
SDS-PAGE comparison of HFbp and NFbp. 5
µg of each protein were run on a 12% acrylamide gel as specified
under ``Experimental Procedures.'' Numbers on the left refer to molecular weights estimated from a reference
curve of standard protein relative
mobilities.
A hitC deletion mutant was constructed in order to
demonstrate the essential nature of this gene to the complementation of
H-1443 to growth on nutrient agar containing 200 µM dipyridyl (NA/Dip
Biochemical analyses of purified
HFbp were performed to compare its physical and functional properties
with those of NFbp. The biochemical attributes of NFbp have been
extensively reported (17, 32, 33, 36) and are listed as
part of Table 2. Physical comparison of HFbp with NFbp reveals
that both proteins share similar predicted molecular masses, although
their migration in SDS-PAGE is noticeably different (Table 2, Fig. 2). The isoelectric points of HFbp and NFbp differ by more
than a full pH unit. This difference in charge may explain the
disparity in SDS-PAGE mobility and affinity for ion exchange resins.
However, two other functional indices highlight the similarity that the
proteins share in their coordination of iron. The visible absorbance
maximum of the ferrated protein is nearly identical for the two,
indicating that iron is bound within a very similar ligand field in
both HFbp and NFbp. Secondly, the affinities for Fe
Figure 3:
Iron saturation of HFbp by
Fe
An
experimental approach based upon that described for defining the sfuABC operon by Zimmerman et al.(25) was
used to investigate iron transport from the periplasmic space to the
cytoplasm. This approach employs E. coli H-1443, an aroB
Figure 4:
Competition for dipyridyl-bound labeled
iron by HFbp in the periplasm. E. coli strain H-1443, with and
without the HFbp-producing plasmid pBSJ1, was grown overnight on
nutrient agar containing 75 µM dipyridyl and
Based upon the above
observations, we investigated the ability of the hitABC operon
to confer growth upon aroB
Figure 5:
Outline of the construction of the hitABC-containing plasmid pAHIO. As described under
``Experimental Procedures,'' a PCR fragment containing the hitABC operon with minimal flanking sequence was amplified
from pJDS150. Using SmaI and BamHI ends, the fragment
was cloned into the EcoRV and BamHI sites of
pBR322.
Figure 6:
Complementation of aroB
The hitABC iron acquisition operon appears to behave as a
classical active transport system dependent upon a periplasmic binding
protein, a cytoplasmic permease, and a nucleotide-binding protein.
Therefore, disruption of any of these three components should eliminate
the function of the system. In order to demonstrate the requirement for
a functional hitABC operon for growth, a 400-bp C-terminal
deletion of the hitC gene, designated pAHIO Complementation for
growth is a gross measure of the molecular events contributed by the hitABC operon. By examining the distribution of labeled iron
in the cell, the efficiency of iron transport at the molecular level
can be measured. We investigated the movement of iron from the
periplasm into the cytosol using
Figure 7:
Comparison of labeled iron uptake by E. coli strain H-1443 isogenic variants containing one of
three plasmids: pBSJ1 (HFbp producing), pAHIO (complete hit operon), or pAHIO
This study represents the characterization of a periplasmic
free iron transport system at the genetic and biochemical levels. By
cloning, overexpressing, and purifying the hitA gene product,
we demonstrated that HFbp competes for iron bound to dipyridyl in
vitro. Furthermore, HFbp can compete in vivo for
dipyridyl-bound iron in the E. coli periplasm. Finally, growth
on dipyridyl-containing agar can be conferred upon the E. coli host by expression of a functional operon. The model proposed in Fig. 8illustrates the mechanism by which dipyridyl-bound iron at
sufficient concentrations diffuses across a permeable outer membrane
into the periplasm, where HFbp is able to compete for this iron source.
Subsequent transport of iron across the cytoplasmic membrane can only
be accomplished in the presence of a functional cytoplasmic membrane
permease and a nucleotide-binding protein, as evidenced by growth as
pin-point colonies on NA/Dip
Figure 8:
Model for hitABC-mediated iron
acquisition from dipyridyl in aroB
Common themes in
the transport of molecules across membranes have begun to emerge from
diverse and detailed studies of prokaryotic and eukaryotic systems. For
example, we have recently demonstrated that NFbp functions in the
transport of iron across the periplasm in a manner analogous to the
transport of iron in serum by the vertebrate Transferrins(35) .
This analogy holds not only at the level of function, but also at the
level of structure: an identical set of iron-binding ligands is used by
NFbp and by transferrin(35) . As a closely related protein
homologue, HFbp can be reasonably predicted to function in a similar
capacity. Another common theme is the utilization of an ABC (for ATP
Binding Cassette) transporter-exporter protein
complex(40, 43) . ABC transporter-exporters require
energy for transport. This energy is supplied by hydrolysis of a
nucleotide triphosphate, which is facilitated by the nucleotide-binding
protein. In eukaryotic systems (e.g. the mammalian
P-glycoprotein drug exporter, MDR), ATP hydrolysis and membrane
permease activities are contained on a single polypeptide. In contrast,
the cytoplasmic permease and nucleotide binding activities exist on two
separate polypeptides in the bacterial ABC
importers(40, 41, 42) . This is consistent
with the data presented above for the transport of growth-essential
iron by the hitABC operon. An important aspect of these
studies is the correlation of this operon family with bacterial
pathogens. The ability to optimally compete for iron from the host
environment correlates with pathogenicity of Neisseria species(44, 45) . The microbial pathogens N.
gonorrhoeae and N. meningitidis both express
antigenically detectable levels of NFbp when propagated under
conditions of iron stress, whereas the closely related commensal Neisseria species (e.g. N. sicca and N.
perflava) do not(32, 36) . Since we have shown
that HFbp requires a functional permease and nucleotide-binding protein
to function, it is presumed that these are also present in pathogenic Neisseria but not in commensal Neisseria. Furthermore, the studies of Sanders et al.(24) demonstrated that a functional hitABC operon
was associated with a non-typable H. influenzae isolate. This
functional operon was used to complement a type b H. influenzae to growth on iron-limited medium. H. influenzae causes a
spectrum of disease, ranging from asymptomatic colonization to invasive
bacterial meningitis(46) . It is possible that the presence of
a functional hit operon may contribute to the differences in
pathogenicity of individual H. influenzae isolates. In
addition to the periplasm-to-cytoplasm transport of iron investigated
in this report, it has been demonstrated that delivering iron to the
periplasmic space is a critical step in obtaining iron from the host
environment. We have recently defined a human transferrin-binding
protein complex oriented at the surface of H.
influenzae(16) . This complex is required as the first
step in the process of assimilation of iron from human transferrin. In
the current study we circumvent the delivery of iron from transferrin
into the periplasm by substituting dipyridyl as an iron source to a E. coli host lacking a functional transferrin receptor
complex. Two molecules of dipyridyl bind a single molecule of iron at
physiologic pH, giving the complex a molecular mass of less than 400
Da, a size that is freely diffusible through the E. coli porin. Alternatively, since dipyridyl is a hydrophobic compound it
may diffuse through the outer membrane by partitioning into the lipid
bilayer in a manner similar to what is observed for
erythromycin(47) . In either case iron is delivered to the
periplasm and the requirement for a functional outer membrane receptor
complex is subverted. These studies demonstrate that the hitABC operon can function independent of the transferrin
receptor. This correlates with the fact that the sfuABC system
from S. marcescens has been demonstrated to be TonB
independent, a characteristic which would preclude the involvement of
an energy-dependent outer membrane receptor complex. The significance
of this mechanism of iron uptake in such a diverse range of
microorganisms and its role in pathogenicity is unknown. HFbp and
NFbp are homologous in their biochemical attributes. Because of this
relatedness, it is likely that both proteins serve similar functions in
these two bacterial species. It has been previously demonstrated that
NFbp transiently associates with labeled iron mobilized from
transferrin. This evidence has implicated NFbp in the transport of iron
from the transferrin-binding complex to the cytoplasm (17) .
This study illustrates that, similar to the sfuABC system, the hitABC system is capable of functioning in the absence of
Tbp1/2 by acquiring iron directly from sources which diffuse into the
periplasm. As a result, the hitABC transporter and the
Neisserial equivalent may function in a broader role as a high
affinity, periplasmic iron scavenging system. Further characterization
of this system will enhance our understanding of this mechanism of iron
acquisition and allow a more detailed molecular understanding of how
the Fbp family of proteins participates in the process of periplasmic
iron transport.
Volume 270,
Number 42,
Issue of October 20, 1995 pp. 25142-25149
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
)genomic DNA fragment from an isolate proficient for growth
on this medium was identified by this analysis(24) . Essential
for this phenotype was a 4-kb operon composed of three genes: hitA, hitB, and hitC (hit for Haemophilus
Iron Transport) proposed to encode a periplasmic iron-binding protein,
a cytoplasmic permease, and a nucleotide-binding protein, respectively.
A homologous three-gene operon was originally described for Serratia marcescens and designated sfu for Serratia Ferric-iron Uptake(25) . The sfu operon was isolated based upon its ability to complement an E.
coli strain (H-1443) deficient in its ability to produce
siderophores for growth on nutrient agar containing 200 µM 2,2`-dipyridyl (dipyridyl), an iron chelator that sequesters free
iron in the medium (25, 26) . The open reading frames
encoded by the hitABC and sfuABC genes were found to
share 38, 37, and 38% identity between respective A, B, and C
components at the predicted amino acid level. The similarities between
the hit and sfu genetic loci suggest a high level of
conservation among two diverse species.
40% identity) with the Neisserial Fbp(17) . Fbp is a periplasmic iron-binding
protein expressed by all pathogenic Neisseria that functions
as the binding component of a high affinity active transport system for
the assimilation of growth-essential iron from the Transferrins.
Purified Fbp binds a single Fe
ion with an affinity
approaching that of the Transferrins (17, 34, 35) and by a mechanism that is
remarkably conserved among this family of proteins, coordinating iron
through two tyrosines, a single histidine, and a bicarbonate
anion(35) . In our study we will refer to these Fbp homologues
as NFbp for Neisseria Fbp derived from the fbp gene
locus (33) , HFbp for Haemophilus Fbp derived from the hitA locus(24) , and SFbp for Serratia Fbp
derived from the sfuA locus(26) . It is clear is that
a common free iron active transport system exists among pathogenic
members of the diverse microbial families Enterobacteriaceae (sfu operon), Pasteurellaceae (hit operon), and Neisseriaceae (fbp operon). The
existence of this common system may reflect its contribution to the
pathogenicity of these organisms.
Strains and General Reagents
Type b H.
influenzae strain DL63 was obtained from E. Hansen, University of
Texas Southwestern Medical Center, Dallas, Texas. E. coli strains DH5
and JM109 were purchased from Promega (Madison,
WI). E. coli strain H-1443 was kindly provided by Dr. Volkmar
Braun, Universität Tübingen,
Tübingen, Germany.
, tetracycline, and
ampicillin were purchased from Sigma. Low molecular weight protein
standards for SDS-PAGE analysis were purchased from Pharmacia Biotech
Inc. The radioisotopes [-P] and
[
Fe
](NO
)
were purchased from DuPont NEN. Eco-Lite scintillation mixture was
purchased from ICN Biomedicals Inc. (Irvine, CA), and samples were
counted using a Packard 1600TR Tri-Carb liquid scintillation analyzer
(Packard Instrument Company, Meridian, CT). Whatman no. 4 filter paper
was purchased from Whatman (Maidstone, United Kingdom). The Amicon
concentration cell and Diaflo ultrafiltration membranes were from
Amicon (Lexington, MA).Cloning of the hitA Gene Region and the Minimal hitABC
Operon
A 700-bp hitA gene fragment was PCR amplified
from H. influenzae DL63 chromosomal DNA by designing a primer
based on the N-terminal amino acid sequence of a 40-kDa iron-regulated
periplasmic protein suspected to be the H. influenzae Fbp
analog (27) and by designing a second primer to a conserved
region shared by the closely related sfuA(26) and
gonococcal fbp genes(33) . PCR reactions were
performed in 50 µl volumes by methods previously
described(28) . The PCR profile consisted of 25 cycles of 94
°C for 1 min, 50 °C for 2 min, and 72 °C for 2.5 min which
was followed by a final extension of 72 °C for 10 min. Cloning of
an intact hitA containing construct was achieved by screening
a previously prepared Zap II H. influenzae DL63
chromosomal DNA library using the
-P 700-bp random
hexamer-labeled hitA PCR product as a probe. The
Zap II
gene library was prepared by random mechanical shearing of DL63
chromosomal DNA followed by blunt ending the sheared products using
S
nuclease. The blunted products were methylated using EcoRI methylase to protect genomic EcoRI sites and
then treated with Klenow to ensure blunt ending of the genomic DNA
fragments. The blunted, methylated fragments were ligated to EcoRI linkers using T4 DNA ligase and then digested with EcoRI restriction endonuclease. The DNA was then
size-fractionated on a sucrose density gradient to obtain 5-10-kb
fragments which were subsequently cloned into the unique EcoRI
site within the lacZ gene pBluescript portion of the Zap
II vector. Plating of the library was done in accordance with the
manufacturer's directions except for the following modifications:
LB broth containing 0.7% Difco agar and 10 mM MgS0
was used in the place of NZY agar; similarly, LB top agar was
used to replace NZY top agar. Plaque lifts were performed on the plated
library by applying nitrocellulose filters to plates prechilled to 4
°C and incubating the plates with the applied filters at this
temperature for 30 min. Subsequently, filters were marked for
orientation, removed from the plates, and screened according to the
Stratagene protocol. After two rounds of screening, a number of
positive clones were identified, and the pBluescript SKII phagemids were excised as per the manufacturer's
instructions. One of these clones was designated pBSJ1 and was shown to
contain an
3.5-kb genomic DNA fragment which included the intact hitA gene, approximately two-thirds of the coding region for hitB, and 1.3 kb of noncoding sequence upstream of hitA (Fig. 1).
3.5-kb fragment
encoding 1.3 kb upstream and 1.2 kb downstream of the HFbp coding
sequence was excised from a positive
Zap II clone and inserted
into the EcoRI site of the plasmid pBS SK
. Ori = origin of replication, AmpR =
-lactamase gene, lacZ = -galactosidase
gene.
250 bp
upstream and 230 bp downstream of this operon. For the upstream primer,
hitO-5`, there was an engineered 5` SmaI site; the downstream
primer hitO-3` included a 3` BamHI site (Table 1). Using
these primers and the plasmid pJDS150 as template, a PCR fragment of
approximately 4.2 kb was generated. Following PCR, the amplified
fragment was gel purified and digested simultaneously with BamHI and SmaI for about 4 h at 37 °C. The PCR
fragment was combined with the 4.2-kb EcoRV-BamHI
fragment of pBR322 (gel-purified) at a 3:1 ratio of insert to vector.
Ligation was achieved using standard conditions(33) . This
ligation was used to transform competent E. coli strain H-1443
and the transformants selected on LB agar containing 100 µg/ml
ampicillin. Transformants were screened for tetracycline sensitivity on
LB plates containing 25 µg/ml of this antibiotic. Tetracycline
sensitive clones were screened for plasmid DNA and the presence of hitABC insert verified by PCR amplification as described
above.
). A 1.3-kb fragment of DNA was
deleted from the ClaI site (
400 bp from the stop codon in hitC) to the NarI site (at position 1205 in pBR322).
This was achieved by complete digestion of pAHIO with NarI
followed by a partial ClaI digest. From this partial digest
the approximately 7-kb partial product that contains the deleted hitC gene was gel-purified. Subsequently, this fragment was
ligated under standard conditions and used to transform E. coli H-1443 cells to ampicillin resistance. Positive clones were
confirmed by restriction digest analysis and the plasmid expressing
this mutation designated pAHIO
hitC.Isolation of HFbp
Similar to what was observed for
the overexpression of NFbp in an E. coli background(32) , overnight growth of JM109(pBSJ1) resulted
in a distinctly red pellet upon centrifugation of a 1.5 ml suspension,
suggestive of the overexpression of HFbp. This was confirmed by
SDS-PAGE analysis of crude CTAB lysates from E. coli JM109
(pBSJ1), prepared as described previously(32) . This analysis
indicated a major protein with a molecular mass of 40 kDa (data
not shown). Purification of HFbp was achieved by a modification of the
method of Berish et al.(32) using 2-liter cultures of
JM109(pBSJ1) in LB supplemented with 200 µg/ml ampicillin, grown
with aeration at 37 °C for 22-24 h. The cells were harvested
by pelleting at 5000
g for 15 min at 4 °C and
washed once in phosphate-buffered saline followed by suspension in 25
ml of 1 M Tris, pH 8.0. Lysis of the cells was performed by
treating this cell suspension with 25 ml of a 4% (w/v) CTAB solution
accompanied by shaking for 1 h at 37 °C. The solution was then
diluted to 400 ml with dH
O, the particulates removed by
centrifugation (7000 g for 15 min at 4 °C), and
the solubilized material reserved. The pelleted debris were suspended
in 25 ml of 1 M Tris, pH 8.0, and subjected to a second round
of CTAB lysis and pelleting of particulate material as described above.
The lysates from the two CTAB lysis steps were then combined and
diluted to a final volume of 1 liter with dHO and clarified
by filtration through Whatman no. 4 filter paper. The clarified lysate
was applied to a CM-Sepharose CL-6B column connected in series to a
DEAE-Sephacel column (6 cm diameter 15 cm length and 6 cm
diameter 7 cm length, respectively). Equilibration of the
CM-Sepharose column was accomplished by washing with five volumes of 1 M NaOH followed by five volumes of 10 mM Tris base,
pH 8.0, containing 1 M NaCl (high salt buffer) and five
volumes of 10 mM Tris base, pH 8.0 (low salt buffer). The DEAE
column was equilibrated by washing with five volumes of high salt
buffer followed by five volumes of low salt buffer. The soluble CTAB
extract was applied to the equilibrated system at a flow rate of 1.5
ml/min and the HFbp-containing eluant collected. This eluant, which
contained >95% HFbp, was concentrated 10-fold in an Amicon cell
using a 10-kDa cutoff Diaflo ultrafiltration membrane. Alternatively,
the protein was precipitated by bringing the eluant to a final
concentration of 80% ethanol (v/v) to yield a pure HFbp precipitate.Biochemical Characterization of HFbp
The
purification of NFbp was performed as described
previously(32) . Protein determinations were obtained using a
modified Lowry method(32) . HFbp and NFbp preparations (5
µg each) were analyzed on a 12% acrylamide gel using reducing
SDS-PAGE conditions as described previously(32) . Molecular
mass estimates were obtained from this analysis using least-squares
method from molecular weight standards on the same gel. Predicted
molecular masses of HFbp and NFbp were obtained from their previously
published DNA sequences(24, 33) . Isoelectric focusing
was performed as described previously for the gonococcal
Fbp(32) . Visible absorbance spectra were determined from a
2-mg/ml sample of purified HFbp in 20 mM Tris, pH 8.0,
containing 200 mM NaCl, as described previously(35) .
HFbp iron affinity was estimated from citrate competition assays
performed using a method nearly identical to that described for NFbp by
Chen et al.(17) . For this analysis, aliquots of HFbp
were incubated with increasing concentrations of citrate, pH 8.0, and
deferration of HFbp was monitored by decrease in absorbance at 483 nm.Partitioning of Labeled Iron between the Periplasm and
Non-periplasmic Compartments of E. coli Expressing hit
Constructs
Bacterial cultures for these assays were prepared by
inoculation with a single colony in 5 ml of nutrient broth containing
100 µg/ml ampicillin (if required) and grown for 8 h at 37 °C.
A 100-µl aliquot of the culture was plated onto nutrient agar
containing 75 µM dipyridyl, 100 µg/ml ampicillin (if
required), and 10
counts/min
[Fe](NO
) (0.1 µmol).
Plates were grown at 37 °C for 12 h and the bacteria harvested. The
bacteria were then washed three times in phosphate-buffered saline,
suspended to an OD of 0.68 in phosphate-buffered saline,
and 1.5 ml of each culture was pelleted. The periplasmic fraction was
isolated using a modification of the method of Ames(22) .
Briefly, pellets were resuspended in 20 µl of chloroform, vortexed,
and incubated at 25 °C for 15 min. 100 µl of 10 mM Tris, pH 8.0, was added to each sample, followed by vortexing.
Samples were pelleted by centrifugation for 5 min at 7000
g and the aqueous phase containing the periplasmic fraction was
removed. The remaining chloroform suspension represents the
non-periplasmic fraction of the bacteria. Samples were counted after
dissolving them in 3 ml of scintillation mixture.Complementation of aroB
Plating
data were obtained by growing strains to mid-log phase in LB
supplemented with ampicillin (where necessary) and diluted to a
concentration that would allow 100-200 colony forming
units/plate. Under these conditions, the aroB E. coli to
Growth on Dipyridyl-containing Agar by the hit Operon
E. coli strain H-1443 grows on nutrient agar containing
100 µM dipyridyl but not on NA/Dip
. The hitA expressing E. coli H-1443(pBSJ1), hitABC expressing H-1443(pAHIO), and the deletion mutant H-1443
(pAHIO
hitC) were investigated for single-colony growth on
nutrient agar containing 100 µM and 200 µM
dipyridyl. DH5 was used as a positive control (data not shown),
whereas H-1443 and H-1443(pBR322) were used as negative controls.
Growth was scored as positive when pinpoint single colonies occurred
after 20 h at 37 °C and negative if no isolated colonies were
observed after 20 h.
Cloning and Overexpression of Recombinant
HFbp
Cloning and sequencing of the hitABC operon were
recently reported by Sanders et al.(24) .
Independently, we have cloned the hitA region gene. As
described under ``Experimental Procedures,'' the gene
encoding HFbp was cloned based upon amino acid sequence homology
between NFbp (33) and SFbp(26) . To accomplish this,
PCR was used to amplify a 700-bp fragment using a primer based on known
N-terminal amino acid sequence of Hfbp (27) (F3, Table 1)
and a second primer (F6, Table 1) which is based on conserved
sequences between NFbp (33) and SFbp(26) . After
sequencing to confirm that it encoded an Fbp homologue, this fragment
was labeled and used to screen a ZapII library of H.
influenzae DL63 DNA as described under ``Experimental
Procedures.'' One positive recombinant phage was subjected to the in vivo excision protocol, allowing for isolation of the
recombinant plasmid designated pBSJ1. This plasmid (Fig. 1) was
used for the overexpression of HFbp (``Experimental
Procedures''). Sequence comparison of the pBSJ1 insert with the
published hitABC operon indicates that this construct contains
the entire coding region for hitA and two-thirds of the hitB open reading frame in addition to
1.3 kb of DNA
upstream of hitA (data not shown).
Purification and Biochemical Characterization of
HFbp
The isolation of HFbp was performed using a modification of
the NFbp purification procedure of Berish et al.(33) .
Cetyltrimethylammonium bromide-solubilized JM109(pBSJ1) cell extracts
were applied to tandemly arranged CM-Sepharose and DEAE-Sephacel
columns. In contrast to the purification of NFbp, in which NFbp binds
tightly to CM-Sepharose, HFbp binds to neither the cation nor the anion
exchanger. While most solubilized proteins were retained on the
columns, HFbp remained in the eluant (data not shown). HFbp in the
eluant was concentrated by ultrafiltration, using an Amicon PM10 filter
with a molecular mass cutoff of 10 kDa. Yields of recombinant HFbp by
this procedure were consistently between 60-90 mg/liter of cell
culture harvested and were of greater than 95% purity as judged by
SDS-PAGE analysis (Fig. 2). are identical, again emphasizing the functional homology between
the two proteins. These and other (37) observations provide
compelling evidence for the functional homology of NFbp and HFbp.
Periplasmic Acquisition of Iron from
Ferric-2,2`-dipyridyl
As the objective of this study was to
demonstrate that NFbp and HFbp function similarly, the ability of these
deferrated purified protein preparations to obtain iron from
Fe-dipyridyl was investigated. Dipyridyl is a
well-characterized organic iron chelator(25) . Fig. 3illustrates that increasing concentrations of
Fe
-dipyridyl, when added to apoHFbp, cause increasing
saturation of HFbp with iron. At high concentrations of
Fe
-dipyridyl, HFbp was completely saturated. These
data demonstrate the ability of HFbp to compete for iron bound to
dipyridyl in vitro and suggest that it should be possible for
HFbp to compete for iron bound to dipyridyl in the periplasm.
(dipyridyl)
. Increasing amounts of
Fe(dipyridyl)
were added to 60 µM apo-HFbp as described under ``Experimental Procedures.''
Binding of iron by HFbp was monitored by the increase in absorbance at
483 nm, the visible maximum of the ferrated protein. This data
demonstrates that apoHFbp can efficiently compete for dipyridyl-bound
iron in vitro.
strain which is deficient in the
synthesis of aromatic compounds (38) including amino acids and
the siderophore enterochelin(39) . The growth of this strain is
inhibited by 200 µM dipyridyl in nutrient agar; however,
at concentrations of 100 µM dipyridyl H-1443 will grow,
presumably due to low affinity iron uptake systems (Table 3). In
order to examine the in vivo competition for dipyridyl-bound
iron by HFbp, H-1443 was grown under conditions in which trace
concentrations of
Fe
-dipyridyl were
incorporated into 75 µM dipyridyl-containing nutrient
agar. This strain and an HFbp-expressing isogenic variant containing
the plasmid pBSJ1 were propagated for 12 h. At this time organisms were
harvested and washed, and periplasmic fractions were extracted from
cells as described under ``Experimental Procedures.'' The
concentrations of labeled iron associated with the periplasm and the
non-periplasmic components were compared for both strains (Fig. 4). The results demonstrate that both strains had
equivalent levels of radioactivity associated with the non-periplasmic
fraction. This is consistent with the observation that these bacteria
share common low affinity systems for iron uptake. In contrast, the
strain expressing HFbp contained 25-fold more
Fe in the
periplasm than did the plasmid-free H-1443. This is consistent with the
prediction that the presence of HFbp in the periplasm would allow
accumulation of free iron from Fe
-dipyridyl at this
site. This demonstrates that, similar to the in vitro ability
of apoHFbp to mobilize iron bound to dipyridyl, periplasmic Hfbp can
effectively liberate Fe
from dipyridyl.
Fe, as specified under ``Experimental
Procedures.'' Bacteria were scraped from plates, washed, and
separated into periplasmic and non-periplasmic fractions. The amount of
iron in either fraction was determined by scintillation counting. The
results demonstrate the ability of the strain expressing HFbp to
efficiently concentrate iron in the
periplasm.
Complementation of a Siderophore-deficient Strain of E.
coli for Growth on Medium containing Ferric-dipyridyl
Production
of a functional siderophore by aroBE.
coli strain DH5
permits its growth on nutrient agar
containing 200 µM dipyridyl (NA/Dip) (Table 3). In contrast, the aroB
E. coli strain H-1443 and the HFbp-expressing pBSJ1
transformant could not be propagated on NA/Dip
but could
be grown on nutrient agar containing 100 µM dipyridyl (Table 3). Thus, the presence of the HFbp is not sufficient for
the removal and transport of iron into the cytoplasm. This implies that
a functional operon is required for complementation, as can be inferred
from the original studies of the sfuABC operon demonstrating
that the entire operon was required for complementation to growth on
this medium (25, 39) .
H-1443 on
NA/Dip
. The plasmid pJDS150 derived by Hansen and
colleagues (24) contains an H. influenzae DNA fragment
encoding the 4.0-kb hitABC operon and an additional 7.5 kb
outside of this operon (Table 1, ``Experimental
Procedures''). This plasmid conferred upon E. coli strain
H-1443 the ability to grow as small colonies on NA/Dip
(data not shown). To demonstrate that only the hitABC genes were required for this functional complementation, a 4.2-kb
PCR fragment containing only hitABC was amplified from pJDS150
and cloned into pBR322. The resulting plasmid pAHIO (Fig. 5)
allowed H-1443 to grow as single colonies on NA/Dip
(Fig. 6, Table 3). Isolated colonies of
H-1443(pJDS150) (data not shown) or H-1443 (pAHIO) were approximately
3-fold smaller than those obtained from plating the aroB
DH5
(data not shown). The negative
controls, untransformed H-1443 (data not shown) and H-1443(pBR322) (Fig. 6), were incapable of growth on NA/Dip.
E. coli for growth on NA/Dip. E. coli strain H-1443 containing either pAHIO or pBR322 were
grown to mid-log and plated to obtain 100-200 colony-forming
units/plate, on either NA or NA/Dip
. Plates A and C represent the growth of H-1443(pBR322) and
H-1443(pAHIO), respectively, on NA/Dip
, while plates
B and D represent their growth on NA. Pinpoint colonies
observed in plate C demonstrate the ability of the hitABC operon to complement H-1443 to growth on dipyridyl-containing
media.
hitC,
was constructed (``Experimental Procedures''). Consistent
with the above prediction, H-1443(pAHIOhitC) was unable
to grow on NA/Dip (Table 3).
Fe to correlate transport
with the presence or absence of the hitABC components. Using
isogenic variants expressing all or part of the operon, the
distribution of
Fe into the periplasmic or the
non-periplasmic cell-associated fractions was analyzed. As described
above, bacteria were grown on
Fe-dipyridyl/nutrient agar
media, harvested, and separated into the periplasmic and the
non-periplasmic cell-associated fractions. Fig. 7shows the
distribution of
Fe for each of the isogenic variants. The
HFbp-expressing H-1443(pBSJ1) showed a level of radioactivity in the
periplasmic fraction similar to that observed for the hitABC operon-expressing H-1443(pAHIO). This is consistent with the
previous assertion that HFbp is effectively mobilizing
Fe
from dipyridyl and concentrating it in the periplasmic space of these
bacteria. However, the greatly increased amount of non-periplasmic
cell-associated radioactivity in H-1443(pAHIO) compared with
H-1443(pBSJ1) indicates that the effective transport of periplasmic
Fe is dependent on the complete hitABC operon
encoded by pAHIO. The profile of the H-1443(pAHIO
hitC)
bacteria is nearly identical to that of the H-1443 (pBSJ1) strain. This
is consistent with the fact that pAHIOhitC should produce
comparable amounts of HFbp but be unable to mediate transport of iron
due to the nucleotide-binding protein deletion.
hitC (complete hit operon
with deletion in the hitC gene). These analogs were grown
overnight on NA containing 75 µM dipyridyl and Fe as specified under ``Experimental
Procedures.'' The isogenic bacterial cultures were scraped from
plates, washed, and separated into periplasmic and cell pellet
fractions. The amount of iron in either fraction was determined by
scintillation counting. These results demonstrate that an intact hitABC operon is required for the transport of iron from the
periplasm to the cytosol, forming the basis for the functional
complementation observed in Fig. 6.
(Fig. 6) and the
increased non-periplasmic cell-associated concentration of labeled iron (Fig. 7). The observation that these colonies were small
relative to those produced by aroB
E.
coli indicates that dipyridyl-associated iron is limiting the
growth of the transformed E. coli under these conditions. The
implication of this observation is that studying the hit operon in this E. coli background may provide an
insightful model for detecting subtle alterations in iron-transport
properties of this operon. In toto, these studies elucidate
the component steps in the active transport of free iron from the
periplasm to the cytosol at the biochemical level.
E. coli.
The E. coli strain H-1443 is unable to access dipyridyl-bound
iron since it lacks the production of a periplasmic transport system.
Expression of HFbp (pBSJ1) allows concentration of iron from dipyridyl
in the periplasm, without further transport to the cytosol (Fig. 7). Supplying the intact hitABC operon (pAHIO)
allows both the concentration of iron in the periplasm and subsequent
transport into the cytosol. As such, this transport system represents
the molecular basis for the growth of pAHIO-containing strains on media
containing the iron chelator dipyridyl.
), nutrient agar
containing 200 µM dipyridyl.
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
ABSTRACT
This article has been cited by other articles:
|
|
 |
|
  K. Doi, Y. Fujino, F. Inagaki, R. Kawatsu, M. Tahara, T. Ohshima, Y. Okaue, T. Yokoyama, S. Iwai, and S. Ogata Stimulation of Expression of a Silica-Induced Protein (Sip) in Thermus thermophilus by Supersaturated Silicic Acid Appl. Envir. Microbiol., April 15, 2009; 75(8): 2406 - 2413. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. R. Mey, E. E. Wyckoff, L. A. Hoover, C. R. Fisher, and S. M. Payne Vibrio cholerae VciB Promotes Iron Uptake via Ferrous Iron Transporters J. Bacteriol., September 1, 2008; 190(17): 5953 - 5962. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Badarau, S. J. Firbank, K. J. Waldron, S. Yanagisawa, N. J. Robinson, M. J. Banfield, and C. Dennison FutA2 Is a Ferric Binding Protein from Synechocystis PCC 6803 J. Biol. Chem., May 2, 2008; 283(18): 12520 - 12527. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. S. Anderson, P. Adhikari, K. D. Weaver, A. L. Crumbliss, and T. A. Mietzner The Haemophilus influenzae hFbpABC Fe3+ Transporter: Analysis of the Membrane Permease and Development of a Gallium-Based Screen for Mutants J. Bacteriol., July 15, 2007; 189(14): 5130 - 5141. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Harrison, W. C. Ray, B. D. Baker, D. W. Armbruster, L. O. Bakaletz, and R. S. Munson Jr. The OxyR Regulon in Nontypeable Haemophilus influenzae J. Bacteriol., February 1, 2007; 189(3): 1004 - 1012. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. W. Whitby, T. M. VanWagoner, T. W. Seale, D. J. Morton, and T. L. Stull Transcriptional Profile of Haemophilus influenzae: Effects of Iron and Heme. J. Bacteriol., August 1, 2006; 188(15): 5640 - 5645. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. A. L. Tom-Yew, D. T. Cui, E. G. Bekker, and M. E. P. Murphy Anion-independent Iron Coordination by the Campylobacter jejuni Ferric Binding Protein J. Biol. Chem., March 11, 2005; 280(10): 9283 - 9290. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. K. Doyle, X. Zhang, V. L. Popov, and J. W. McBride An Immunoreactive 38-Kilodalton Protein of Ehrlichia canis Shares Structural Homology and Iron-Binding Capacity with the Ferric Ion-Binding Protein Family Infect. Immun., January 1, 2005; 73(1): 62 - 69. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. S. Anderson, P. Adhikari, A. J. Nowalk, C. Y. Chen, and T. A. Mietzner The hFbpABC Transporter from Haemophilus influenzae Functions as a Binding-Protein-Dependent ABC Transporter with High Specificity and Affinity for Ferric Iron J. Bacteriol., September 15, 2004; 186(18): 6220 - 6229. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. H. Y. Lau, R. T. A. MacGillivray, and M. E. P. Murphy Characterization of a Nucleotide-Binding Domain Associated with Neisserial Iron Transport J. Bacteriol., May 15, 2004; 186(10): 3266 - 3269. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Dhungana, C. H. Taboy, D. S. Anderson, K. G. Vaughan, P. Aisen, T. A. Mietzner, and A. L. Crumbliss Bioinorganic Chemistry Special Feature: The influence of the synergistic anion on iron chelation by ferric binding protein, a bacterial transferrin PNAS, April 1, 2003; 100(7): 3659 - 3664. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. E. Swords, D. L. Chance, L. A. Cohn, J. Shao, M. A. Apicella, and A. L. Smith Acylation of the Lipooligosaccharide of Haemophilus influenzae and Colonization: an htrB Mutation Diminishes the Colonization of Human Airway Epithelial Cells Infect. Immun., August 1, 2002; 70(8): 4661 - 4668. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. A. Webb, J. W. Moffett, and J. B. Waterbury Iron Stress in Open-Ocean Cyanobacteria (Synechococcus, Trichodesmium, and Crocosphaera spp.): Identification of the IdiA Protein Appl. Envir. Microbiol., December 1, 2001; 67(12): 5444 - 5452. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Katoh, N. Hagino, and T. Ogawa Iron-Binding Activity of FutA1 Subunit of an ABC-type Iron Transporter in the Cyanobacterium Synechocystis sp. Strain PCC 6803 Plant Cell Physiol., August 1, 2001; 42(8): 823 - 827. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. D. Cope, Z. Hrkal, and E. J. Hansen Detection of Phase Variation in Expression of Proteins Involved in Hemoglobin and Hemoglobin-Haptoglobin Binding by Nontypeable Haemophilus influenzae Infect. Immun., July 1, 2000; 68(7): 4092 - 4101. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Dugourd, C. Martin, C. R. Rioux, M. Jacques, and J. Harel Characterization of a Periplasmic ATP-Binding Cassette Iron Import System of Brachyspira (Serpulina) hyodysenteriae J. Bacteriol., November 15, 1999; 181(22): 6948 - 6957. [Abstract] [Full Text]
|
|
|
|
 |
|
 X.-J. Yu, P. A. Crocquet-Valdes, L. C. Cullman, V. L. Popov, and D. H. Walker Comparison of Ehrlichia chaffeensis Recombinant Proteins for Serologic Diagnosis of Human Monocytotropic Ehrlichiosis J. Clin. Microbiol., August 1, 1999; 37(8): 2568 - 2575. [Abstract] [Full Text]
|
|
|
|
 |
|
 L. B. Tabatabai and G. H. Frank Conservation of Expression and N-Terminal Sequences of the Pasteurella haemolytica 31-Kilodalton and Pasteurella trehalosi 29-Kilodalton Periplasmic Iron-Regulated Proteins Clin. Vaccine Immunol., July 1, 1999; 6(4): 617 - 620. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Zhou, W.-D. Hardt, and J. E. Galan Salmonella typhimurium Encodes a Putative Iron Transport System within the Centisome 63 Pathogenicity Island Infect. Immun., April 1, 1999; 67(4): 1974 - 1981. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. H. Khun, S. D. Kirby, and B. C. Lee A Neisseria meningitidis fbpABC Mutant Is Incapable of Using Nonheme Iron for Growth Infect. Immun., May 1, 1998; 66(5): 2330 - 2336. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. W. Bearden, T. M. Staggs, and R. D. Perry An ABC Transporter System of Yersinia pestis Allows Utilization of Chelated Iron by Escherichia coli SAB11 J. Bacteriol., March 1, 1998; 180(5): 1135 - 1147. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| All ASBMB Journals | Molecular and Cellular Proteomics |
| Journal of Lipid Research | ASBMB Today |