|
Originally published In Press as doi:10.1074/jbc.M201950200 on June 28, 2002
J. Biol. Chem., Vol. 277, Issue 37, 34247-34253, September 13, 2002
Pore Formation and Function of Phosphoporin PhoE
of Escherichia coli Are Determined by the Core Sugar
Moiety of Lipopolysaccharide*
Sven O.
Hagge ,
Hans
de Cock§,
Thomas
Gutsmann,
Frank
Beckers§,
Ulrich
Seydel, and
Andre
Wiese¶
From the Research Center Borstel, Center for Medicine
and Biosciences, Department of Immunochemistry and Biochemical
Microbiology, Parkallee 1-40, D-23845 Borstel, Germany and the
§ Department of Molecular Cell Biology, Institute for
Biomembranes, Utrecht University, Padualaan 8, 3584 CH Utrecht, The
Netherlands
Received for publication, February 27, 2002, and in revised form, June 11, 2002
 |
ABSTRACT |
The lipid matrix of the outer membrane of
Gram-negative bacteria is an asymmetric bilayer composed
of a phospholipid inner leaflet and a lipopolysaccharide outer leaflet.
Incorporated into this lipid matrix are, among other macromolecules,
the porins, which have a sieve-like function for the transport or
exclusion of hydrophilic substances. It is known that a reduced amount
of porins is found in the outer membrane of rough mutants as compared with wild-type bacteria. This observation was discussed to be caused by
a reduced number of insertion sites in the former. We performed
electrical measurements on reconstituted planar bilayers composed of lipopolysaccharide on one side and a phospholipid mixture
on the other side using lipopolysaccharide from various rough mutant
strains of Salmonella enterica serovar Minnesota. We found
that pore formation by PhoE trimers that were added to the
phospholipid side of the bilayers increased with the increasing length
of the lipopolysaccharide core sugar moiety. These results allow us to
conclude that the length of the sugar moiety of lipopolysaccharide is
the parameter governing pore formation and that no particular insertion
sites are required. Furthermore, we found that the voltage gating of
the porin channels is strongly dependent on the composition of the
lipid matrix.
| |
INTRODUCTION |
The cell envelope of Gram-negative bacteria consists of the
cytoplasmic membrane, the peptidoglycan layer, and an additional barrier, the outer membrane
(OM),1 (1) which is strictly
asymmetric with respect to its lipid composition. Whereas the inner
membrane (IM) is composed on both sides of phospholipids, the OM
consists of a phospholipid inner leaflet and a lipopolysaccharide (LPS)
outer leaflet. The LPS consists of an oligo- or polysaccharide portion
covalently linked to a lipid component termed lipid A, which anchors
the molecule in the membrane (2). In wild-type strains, the
polysaccharide portion consists of an O-specific chain and
the core oligosaccharide. Rough mutant strains do not express the
O-specific chain, but retain core oligosaccharides of
varying length. The LPS of various rough mutants are characterized by
chemotypes in a sequence of decreasing length of the core sugar as Ra
(complete core), Rb, Rc, Rd, and Re. Deep-rough LPS (Re LPS) represents
the minimal structure of LPS consisting of only lipid A and two
3-deoxy-D-manno-oct-2-ulosonic acid (Kdo)
monosaccharides (3) (Fig. 1).
View larger version (33K):
[in this window]
[in a new window]
|
Fig. 1.
Chemical structures of the LPS used for outer
membrane reconstitution. A, chemical structure of Re
LPS from various Gram-negative species, i.e. E. coli strain F515, S. minnesota strain R595, and
P. mirabilis strain R45. B, schematic chemical
structure of the core region of LPS from different rough mutant strains
of S. minnesota and its chemotypes (Ra to
Re). The phosphate residue at the second heptose and the
2-aminoethyl diphosphate residue at the first heptose are only present
for the mutants R345 and R60.
|
|
The OM protects the cell from harmful agents like antibiotics and
toxins and against changes in osmotic pressure. Transmembrane proteins,
the porins, allowing the uptake and disposal of small hydrophilic
compounds such as nutrients and waste products (4), are assembled in
the OM. In Escherichia coli OmpF and OmpC represent the
general diffusion pores. The phosphoporin PhoE is synthesized when
cells are grown under phosphate limitation (5). PhoE has a molecular
weight of Mr 36,822 and an exclusion size of
Mr ~600 and is weakly anion selective (6). The
crystal structure of PhoE has been solved (7). The channel-forming
motif of PhoE is a 16-strand anti-parallel  -barrel. Short
-hairpin turns define the periplasmic end of the barrel, whereas
long irregular loops are found at the cell surface. The functional unit
of this protein is a trimer.
Outer membrane proteins are synthesized as precursor proteins in the
cytoplasm. After their translocation across the IM via the Sec
machinery (8) and processing to mature protein, they are assembled into
the OM. The amino-terminal signal sequence is removed during or after
translocation through the IM (9). It has been shown that Re mutants
contain reduced amounts of PhoE (10) and other outer membrane proteins
(11). From in vitro studies it is known that the efficiency
of assembly of porins in the OM decreases from strains with an LPS with
a complete core sugar moiety to those with an LPS with a short sugar
moiety. For PhoE a decrease of 95% was found for a Re mutant in
comparison to a Ra mutant (12). Furthermore, it has been shown in an
in vitro model that the chemical structure and the
biophysical properties of LPS are important for a correct and efficient
folding of PhoE into a native-like, trimerization-competent, folded
monomeric form (13). The folding of in vitro synthesized
non-native PhoE protein into a native-like monomeric form was more
efficient with LPS of chemotype Ra (Ra LPS) as compared to Re LPS,
lacking a large part of the core. The absence or presence of a negative charge at the first heptose in the core region seemed to be of special
importance in this respect. Also, the efficiency of folding and
trimerization of OmpF porin in vitro was shown to depend on the LPS structure (14).
In the present study, we examine the effect of the chemical structure
of LPS, in particular that of the length of the core sugar moiety on
the pore formation by PhoE trimers and its pore function in the OM. To
this end, electrical measurements were performed on reconstituted OM as
asymmetric planar membranes according to the Montal-Mueller technique.
For porin from Paracoccus denitrificans, the role of LPS in
pore formation and pore function has been shown earlier (15, 16). These
authors prepared membranes from Re LPS from S. enterica
serovar Minnesota (S. minnesota) strain R595 or
monosaccharide-type fraction of glycosphingolipid (GSL-1), which
substitutes for LPS in the OM of Sphingomonas paucimobilis (17), on one side and a phospholipid mixture (PL) resembling the
composition of the inner leaflet on the other side. It was found that
the pore formation rate in glycolipid/PL membranes was by an order of
magnitude higher than in symmetric PL membranes. The influence of the
length of the saccharide portion was, however, not investigated.
Therefore, we then focused on the effect of the core oligosaccharide on
pore formation rate of PhoE and on the voltage-dependent
gating of the pores formed. Instead of LPS from E. coli,
that from S. minnesota, which has a high homology to
LPS from E. coli, was used because of the availability of a series of LPS with variable length of the sugar moiety (Fig.
1B).
We found that in general the pore formation rate increased with
increasing length of the sugar moiety, whereas only slight differences
in the pore formation rate could be observed for Re LPS from various
strains. The voltage-dependent gating of PhoE pores was
strongly dependent on the composition of the lipid matrix. In general,
the voltage necessary for closure of PhoE was |VC|  50 mV. Therefore, the transmembrane voltage across the bacterial OM of
VOM = 26 mV (inside negative) (18) seems to be too low to
initiate voltage-dependent closure of PhoE; however, the
transmembrane voltage across the IM of VIM = 140-160 mV
(inside negative) (19) would lead to closure of accidentally
incorporated porins in the IM. Thus, porin gating could be an important
mechanism for self-protection of Gram-negative bacteria.
| |
EXPERIMENTAL PROCEDURES |
Lipids and Other Chemicals--
Rough mutant LPS from S. minnesota strains R595 (LPS R595), R7 (LPS R7), R5 (LPS R5), R345
(LPS R345), R60 (LPS R60), E. coli strain F515 (LPS F515),
and P. mirabilis strain R45 (LPS R45) were used in
experiments with asymmetric PL/LPS membranes. The chemical structure of
various Re LPS used in our experiments is shown in Fig. 1A,
those of the core region of different chemotypes of LPS from S. minnesota in Fig. 1B. LPS was extracted by the phenol/chloroform/petroleum ether method (20) and purified, lyophilized, and transformed into the triethylamine salt form. The
amount of the nonstoichiometric substitutions was analyzed by
matrix-assisted laser desorption ionization time-of-flight mass
spectrometry. Phosphatidylethanolamine (PE) from E. coli, phosphatidylglycerol (PG) from egg yolk lecithin (sodium salt), and
synthetic diphosphatidylglycerol (DPG) were purchased from Avanti Polar
Lipids (Alabaster, AL) and used without further purification.
LPS-free Phosphoporin PhoE from E. coli--
The mature form of
PhoE protein was expressed and isolated from E. coli strain
Bl21(DE3) cells containing plasmid pCJ2 and refolded into a
trypsin-resistant form as described previously (21). After refolding,
the proteins were treated with trypsin (30 µg/ml) for 20 min at
37 °C followed by an incubation with 1 mM
phenylmethylsulfonyl fluoride at 0 °C for 1 h to inhibit the
protease. After concentration of the protein solution with centriprep
10 (Amicon, Beverly, MA, United States), the solution was dialyzed
against buffer A (10 mM Tris-HCl, pH 8.0, 0.46 mM Tween 20) with 3 changes at 4 °C. Refolded proteins
were further purified by anion exchange chromatography on Q-Sepharose
HR (Amersham Biosciences) using a Gradifrac (Amersham Biosciences), and
bound protein was eluted from the column with a linear gradient of
buffer A to 0.5 M NaCl. Peak fractions containing the
refolded protein were pooled and dialyzed against buffer A at 4 °C.
Aliquots of protein solution (~0.5 mg/ml protein) were rapidly frozen
with liquid nitrogen and stored at  20 °C. Prior to
electrophoresis, samples containing refolded proteins incubated in 2%
SDS-sample buffer and incubated for 10 min at room temperature,
56 °C or 100 °C, were analyzed by SDS-polyacrylamide gel
electrophoresis (SDS-PAGE) as described previously (21). Refolded
protein consisted mostly of heat-stable trimers (Fig.
2, T) with small amounts of folded monomers (Fig. 2, m*, lane 1) and very low
amounts of dimeric forms (Fig. 2, D, hardly visible in
lane 1). Heating for 10 min at 56 °C will denature
monomeric, dimeric, and small amounts of heat-unstable forms of PhoE,
resulting in detection of heat-stable trimers (Fig, 2, T,
lane 2). After heating for 10 min at 100 °C all forms
were denatured, resulting in detection of denatured PhoE monomers (Fig.
2, m, lane 3). PhoE was diluted in aqueous solution containing 0.15 mg/ml Triton X-100 and stored at 4 °C for
at least 3 days before usage.
View larger version (37K):
[in this window]
[in a new window]
|
Fig. 2.
SDS-PAGE of in vitro folded
and purified PhoE. After folding, trypsin treatment, and
purification by anion exchange chromatography, PhoE protein was
analyzed by SDS-PAGE. The protein sample was dissolved in sample buffer
containing 2% SDS and subsequently incubated for 10 min at room
temperature (lane 1), 56 °C (lane 2), or
100 °C (lane 3). Proteins were stained with Coomassie
Brilliant Blue R-250 after electrophoresis. The position of the trimers
(T), dimers (D), denatured monomers
(m), and folded monomers (m*) are
indicated.
|
|
Preparation of Planar Bilayers and Electrical
Measurements--
For the investigation of pore formation by PhoE and
its pore function in different lipid matrices, planar bilayers
according to the Montal- Mueller technique (22) were prepared as
described previously (23). Briefly, asymmetric bilayers were formed by opposing two lipid monolayers, one from LPS, the other from PL, prepared on aqueous subphases (1.5 ml each) from chloroformic solutions
of the lipids at a small aperture (typically 150 µm diameter) in a
thin Teflon septum (12.5 µm thickness). PL is composed of PE, PG, and
DPG in molar ratios of 81:17:2 resembling the phospholipid composition
of the inner leaflet of the outer membrane of S. enterica serovar Typhimurium (24).
For electrical measurements, the planar membranes were voltage clamped
via a pair of Ag/AgCl electrodes connected to the head stage of a
patch-clamp amplifier. In all cases the compartment opposite where the
porin was added was grounded.
If not otherwise mentioned, measurements were performed at a
temperature of 37 °C with subphases consisting of 100 mM
KCl, 5 mM MgCl2, and 5 mM HEPES
adjusted to pH 7.0. The specific electrical conductivity of this
bathing solution at 37 °C was 17.2 millisiemens cm 1.
At the beginning of each experiment the correct membrane formation was
checked by measuring membrane current and capacitance. Only membranes
with a basic current of less than ±2.5 pA at a clamp voltage of ±100
mV were used for the experiments. PhoE was added to the PL side of the bilayer.
| |
RESULTS |
Pore Formation--
To investigate the influence of the lipid
matrix on pore formation, 20 ng PhoE were added to the PL side of
various PL/LPS membranes. Fig. 3
represents a typical current trace for pore formation by PhoE added to
the PL side of an asymmetric lipid bilayer with Ra LPS from S. minnesota strain R60 on the opposite side at a clamp voltage of 20 mV. 90 s after porin addition, the current increased stepwise.
About 6 min after porin addition the pore formation rate declined, and
pore formation stopped about 30 min after porin addition. The number of
pores was calculated from the current traces as the quotient of
membrane and single-channel current. When PhoE (up to an amount of 2 µg) was added to the LPS side of a PL/LPS membrane, no increase in
membrane current was observed. Control experiments with 4.5 µg Triton
X-100 showed no changes in membrane current.
View larger version (15K):
[in this window]
[in a new window]
|
Fig. 3.
Pore formation by PhoE in PL/LPS R60
membranes. A typical current trace at a clamp voltage of 20 mV is
plotted exemplarily for LPS from S. minnesota strain R60.
Approximately 1 to 2 min after the addition of 20 ng PhoE to the PL
side of the membrane, a stepwise increase of membrane current is
observed. Approximately 30 min after porin addition, pore formation
ceases. The PL side consists of 81 mol% PE, 17 mol% PG, and 2 mol%
DPG. The subphase contains 100 mM KCl, 5 mM
MgCl2 and is buffered with 5 mM HEPES. The
mixture is then adjusted to a pH of 7.0 and heated to a
temperature of 37 °C.
|
|
The histogram of amplitudes of current steps is shown in Fig.
4 exemplarily for PL/LPS R60 membranes.
The single-channel current was derived from the histogram by gauss
approximation. Table I summarizes the
mean current increments of single PhoE trimers in different lipid
matrices. For all membranes, nearly no differences in the
single-channel current were observed. Only for lipid matrices containing LPS R5 and PG slightly, but not significantly, smaller single-channel currents were determined.
View larger version (14K):
[in this window]
[in a new window]
|
Fig. 4.
Single-channel current  . The
single-channel currents of PhoE at a clamp voltage of 20 mV in
different lipid matrices were obtained from the current traces. A
typical histogram of single-channel current is plotted exemplarily for
PL/LPS R60 membranes. Composition of PL side and of subphase as in Fig.
3.
|
|
View this table:
[in this window]
[in a new window]
|
Table I
Single-channel Current
The single-channel currents of PhoE at a clamp-voltage of 20 mV in
different lipid matrices composed on the one side of PL and on the
other of different chemotypes of LPS from S. minnesota, Re
LPS from various Gram-negative species, i.e., E. coli strain
F515, P. mirabilis strain R45, and PG as a charged
phospholipid were obtained by gauss-approximation from the histograms
of amplitudes of current steps. The error  is given as 2  . Only the side opposite to porin addition is itemized. The other
side was composed of PL. Composition of PL side and of subphase as in
Fig. 3.
|
|
The mean number of pores formed by PhoE in different lipid matrices
made from PL on the side of porin addition and of rough mutant LPS from
S. minnesota on the opposite side at a clamp voltage of 20 mV at 5, 10, 15, and 20 min after porin addition are presented in Fig.
5A. Each value represents the
mean number of pores found in 3-5 successful experiments. Because of
membrane rupture at an early stage, many experiments could not be
considered. In general, membrane stability decreased with
increasing length of the core sugar moiety. Furthermore, we were not
successful forming stable bilayers using LPS from S. minnesota strains R4. The mean number of pores formed within 20 min decreases from LPS R60, which is set to 100%, over LPS R345
(71%), LPS R5 (16%), and LPS R7 (24%) to LPS R595 (6%). The
sequence of the traces representing the data obtained with the various
LPS chemotypes is the same at any time point, i.e. they do
not cross. According to Student's t test, the significance
level of the number of pores incorporated up to 20 min between the two
most different chemotypes, LPS R60 and LPS R595, is <0.01. The
corresponding significance levels between two neighboring traces
are  0.61 (e.g. 0.61 for LPS R60 and LPS R345, 0.60 for LPS
R345 and LPS R7, 0.48 for LPS R7 and LPS R5, and 0.23 for LPS R5 and
LPS R595) because of the large error bars. The significance levels
between more distant traces are smaller than 0.15, with the exception
of the value between LPS R345 and LPS R5 (0.34).
View larger version (21K):
[in this window]
[in a new window]
|
Fig. 5.
Mean number of pores. The mean
numbers of pores after addition of 20 ng PhoE were investigated
in planar bilayer experiments at a clamp voltage of 20 mV for several
lipid matrices composed of PL on the side of porin addition and on the
other of (A) different chemotypes of LPS from S. minnesota and (B) Re LPS from various Gram-negative
species, i.e. E. coli strain F515, P. mirabilis strain R45, and PG as a charged phospholipid. Each
plotted value represents the mean current of at least three
experiments. Composition of PL side and of subphase as in Fig. 3.
|
|
In Fig. 5B, the mean numbers of PhoE pores (determined as
described before) in lipid matrices composed of PL on the side of porin
addition and Re LPS from various Gram-negative species (i.e. S. minnesota strain R595, E. coli strain F515,
and P. mirabilis strain R45) on the other side are depicted.
PG has been used as an example of a negatively charged phospholipid.
The number of pores at 20 min decreased from 15% in the case of
LPS F515, 9% (LPS R45), and 6% (LPS R595) to, in the case of PG, 2%
of the values found for LPS R60. The value for symmetrical PL/PL
membranes is comparable to those found for PL/PG membranes (data not shown).
Furthermore, it should be mentioned that in general the final number of
pores reached increases from Re LPS to Ra LPS. For LPS with a short
sugar moiety this finish is reached 15 to 20 min after porin addition.
For LPS with a longer sugar moiety, in particular for LPS R60 and LPS
R345, the maximum is reached about 20-30 min after porin addition.
The effect of the Mg2+ concentration on pore formation by
PhoE was determined for PL/LPS F515 membranes prepared in different subphases containing 1, 5, and 15 mM MgCl2,
respectively. The rate of pore formation after the addition of 20 ng
PhoE decreased drastically with increasing Mg2+
concentration (data not shown).
Pore Function--
About 20 min after porin addition, the increase
in membrane current ceased. To characterize the
voltage-dependent closure of PhoE, current/voltage
characteristics of the PhoE trimers were recorded by applying a
triangular voltage (3.2 mV/s) with an amplitude of 200 mV. All curves
were recorded at least twice from independent membrane preparations.
The recordings were repeated several times in each experiment (until
the membrane broke), and the depicted curves in Fig.
6 are in each case averages over all
respective data. The plotted current is normalized to the current
determined at a clamp voltage of +40 mV at which all pores were found
in the open state. At small absolute clamp voltages (|V| < 50 mV) the membrane current was proportional to the transmembrane voltage. At
higher absolute clamp voltages an underproportional increase, a steady
state, or a decrease of membrane current was detected because of the
closure of porin channels. The current/voltage characteristics show a
hysteretic behavior. The efficiency of voltage-dependent
closure can be taken from the onset and the shape of the hysteresis
loops.
View larger version (30K):
[in this window]
[in a new window]
|
Fig. 6.
Voltage-dependent gating of PhoE.
The current/voltage-characteristics of PhoE were determined from
planar bilayer experiments by applying a triangular voltage (3.2 mV/s)
to asymmetric lipid matrices composed of PL on the side of porin
addition and on the other of various LPS chemotypes. All
curves were recorded at least twice from independent
membrane preparations. The recordings were repeated several times in
each experiment, and the depicted curves are in each case
averages over all respective data. The current is normalized to the
current determined at a clamp voltage of +40 mV. The clamp voltages
VC+ and VC are the voltages at which
deviation from linearity in the positive and in the negative voltage
range, respectively, occur because of channel closing. A,
rough mutant LPS from S. minnesota strains R595 (black
trace, VC+ ~80 mV, VC < 200 mV), R7
(cyan trace, VC+ ~65 mV, VC
~150 mV), R5 (green trace, VC+ ~60 mV,
VC ~150 mV), and R345 (red trace,
VC+ ~95 mV, VC ~145 mV). B,
Re LPS from E. coli strain F515 (blue trace,
VC+ ~100 mV, VC ~110 mV), S. minnesota strain R595 (black trace, VC+
~80 mV, VC < 200 mV), P. mirabilis strain
R45 (red trace, VC+ ~80 mV, VC
~120 mV), PG (green trace, VC+ ~110 mV,
VC ~100 mV), and PL (cyan trace,
VC+ ~125 mV, VC ~115 mV). Composition
of PL side and of subphase as in Fig. 3.
|
|
In Fig. 6A current/voltage characteristics of PhoE
incorporated into different lipid bilayers made from PL on the side of porin addition and from rough mutant LPS from various S. minnesota strains on the other side are shown. To better
distinguish between different traces the respective clamp voltages at
which deviation from linearity in the positive and in the negative
voltage ranges occurred because of channel closing. VC+ and
VC are given in the figure legend. At negative clamp
voltages nearly no voltage-dependent closure of the porin
channels occurred. In the positive range at clamp voltages above 50 mV,
a hysteretic current/voltage characteristic was observed. The
efficiency of voltage-dependent closure increased with the
length of the sugar portion of the LPS from LPS R595 over LPS R7 to LPS
R5. In the case of LPS R345, the efficiency of
voltage-dependent closure decreased in comparison to LPS
R5. In the case of LPS R60, we were not able to successfully record a
current/voltage characteristic because of the instability of the lipid bilayer.
In Fig. 6B, the current/voltage characteristics of
PhoE incorporated into lipid bilayers containing Re LPS from various
species (i.e. S. minnesota strain R595, E. coli strain F515, and P. mirabilis strain R45), PG as a
negatively charged phospholipid, and PL (reconstitution of the
cytoplasmic membrane) on the one side and PL on the other side are
depicted. For all LPS, except that from S. minnesota strain
R595, an effective closure of PhoE was also observed at negative clamp
voltages of about 100 to 120 mV. At positive clamp voltages
for LPS F515 a strong decrease in membrane current occurred. For LPS
R45 and LPS R595 only a steady state was reached, and accordingly the
hysteretic character of the current trace was much smaller. For both
phospholipid systems effective closure of PhoE was observed in the
positive voltage range. At negative clamp voltages, gating was
much more pronounced in PL/PG than in PL/PL membranes, yielding almost
symmetric current/voltage characteristics in the former and an
asymmetric characteristic in the latter case.
| |
DISCUSSION |
In the past, several investigations have been published aiming at
an understanding of the role of LPS in the biogenesis of porins.
Performing in vitro folding studies, it has been shown that
LPS is required for correct and efficient folding of PhoE into a folded
monomeric intermediate (13) and that LPS influences folding and
trimerization of OmpF protein (25). In the present investigation we
observed that the composition of the lipid bilayer, and particularly
the chemical structure of LPS used for membrane reconstitution, plays
an important and probably major role in the incorporation and/or late
steps of pore formation by PhoE. Furthermore, we found that the
electrical properties of the trimer are influenced by the composition
of the lipid matrix.
It was shown previously in in vitro studies that the
efficiency of assembly of an in vitro synthesized PhoE
protein into the OM of different E. coli mutants decreases
by about 95% from Ra to Re mutants (12). In that system, folding,
trimerization, and insertion into the OM could not be distinguished.
Subsequently it was shown that LPS was required in an early step in
folding, the formation of the folded monomer (26). Re LPS was far less efficient in supporting the folding of PhoE as compared with Ra LPS
(13). These monomers represent genuine intermediates since they could
be trimerized into native-like trimers after incubation with OM, in
which they appeared to be inserted since they could not be extracted
with urea. Now we observe that the composition of the lipid matrix has
a strong influence on late steps in the biogenesis, insertion, and/or
pore formation of in vitro folded and LPS-free porin
since the efficiency of pore formation by PhoE differs in lipid
matrices containing different chemotypes of LPS from S. minnesota. In general, the number of pores incorporated at a given
time decreased from Ra LPS (LPS R60) to Re LPS (LPS R595) (Fig
5A). The brought distribution of the mean number of pores at
a given time in a particular membrane system can be explained by
differences in the time for diffusion of the porins from the locus of
addition to the subphase to the membrane surface in repetitive experiments. Because of the partly overlapping error bars, a
quantitative correlation between pore formation and lipid composition
can not be established between traces corresponding to neighboring
chemotypes. However, with increasing distance between the chemotypes
the significance levels decrease to p < 0.01 between
LPS R60 and LPS R595. Therefore, from the fact that the sequence in the
number of pores incorporated into the various lipid matrices was the
same at any time, it can be deduced that the lipid matrix determines
the efficiency of porin assembly into the OM. The presence of LPS with
a longer core sugar moiety on the membrane leaflet opposite porin
addition thus promotes all steps in porin assembly.
Now, as the composition of the lipid bilayer, in particular the LPS
chemotype on the side opposite to porin addition, determines the pore
formation rate, the question arises as to which physicochemical membrane properties are responsible for this observation. Here in the
first place membrane fluidity and surface charge density have to be considered.
It has been shown earlier that an increase in membrane fluidity
increases the formation of lytic membrane pores by the complement system (27). However, this parameter cannot explain the positive correlation between pore formation and the length of the core sugar
moiety, because membrane fluidity varies only slightly for the LPS used
in the present investigation (28). Furthermore, the least efficient
pore formation was observed in PL/PG and PL/PL membranes that have the
highest fluidity of all lipids used.
In the case of asymmetric PL/LPS bilayers, a potential gradient arises
from the higher surface charge density on the LPS side as compared with
the PL side. Thus, for the influence of the core sugar moiety on pore
formation, a positive correlation between the number of negative
charges and the pore formation by PhoE can be stated. LPS R60 and LPS
R345, which give rise to the highest pore formation, carry at least 5 negative charges. LPS R7, LPS R5, and LPS F515, with 4 negative charges
each, are in the intermediate range. LPS R595 and LPS R45, carrying
less than 4 charges each, show even smaller pore formation rates. Also,
the decrease in pore formation at higher Mg2+ concentration
could be explained by a reduction of negative charges on the LPS side.
However, it should be mentioned that the higher pore formation rate in
PL/LPS R60 membranes compared with PL/LPS R345 membranes, as well as
the same rate in PL/LPS R595 and PL/LPS R45 membranes, cannot be
explained by differences in the respective LPS charges. Furthermore, it
can be taken from the current/voltage characteristics that a high
external potential does not induce any further pore formation (Fig. 6,
A and B). This effect might be explained assuming
that the potential gradient is not responsible for the differences in
pore formation rates. Rather, specific interactions between the
divalent cations and the negative charges at the outer side of
the porin and the negative groups of the glycolipids stabilizing the
porin in its functional transmembrane configuration (29, 30) are
responsible for the differences in pore formation rates. These specific
interactions might also explain why the number of pores found in PL/LPS
R7 membranes is slightly higher than in PL/LPS R5 membranes, although
the core region of the latter carries an additional glucose.
In addition to membrane fluidity and negative surface charge, several
further structural parameters of LPS that might influence pore
formation are known. Thus, the tendency of the lipid A portion of the
LPS to form nonlamellar structures (31) might influence the stability
of the LPS side of the PL/LPS bilayer. For the S. minnesota
chemotypes, which have identical lipid A portions, membrane stability
might be further reduced by the increasing length of the core sugar
because of the increasing hydrophilicity of the LPS and repulsion
between increasing numbers of negative charges. Within the context of
membrane stability, the reduced pore formation by PhoE in PL/LPS F515
bilayers prepared in subphases containing higher amounts of
Mg2+ can be understood on the basis of an increased
membrane stability attributed to cross linking of LPS molecules by the
divalent cations (4). A reduction of the stability of the LPS layer
might do one of the following: (i) allow pore formation by porins
already inserted into the PL side or (ii) influence the stability of
the entire membrane and thus facilitate insertion and pore formation. The former possibility seems to be unlikely, because lower amounts of
PhoE (10, 32) and other outer membrane proteins (11) have been found in
Re mutants than in wild-type bacteria, indicating that these proteins
do not insert into the membrane.
The differences in pore formation between PL/LPS R595 and PL/LPS R45
membranes on the one and PL/LPS F515 membranes (Fig. 5B) on
the other hand can be explained by the additional
4-amino-4-deoxy--l-arabinose at the lipid A portion that leads to a
reduction of the net negative charge from 4 to 3.5. However, the second
4-amino-4-deoxy--l-arabinose on the first Kdo of LPS R45, which
causes a further reduction of the charge to 3, has no influence on pore
formation. This might be explained by a compensation of the
destabilizing contribution of the higher charge of LPS R595 by the
stabilizing effect of an additional fatty acid (Fig.
1A).
Our findings also answer a question raised by Korteland and Lugtenberg
(32). These authors had shown in in vivo studies that the
uptake of negatively charged substrates by PhoE was more efficient in
Re mutants. They stated that these findings are a consequence of one of
the following: (i) an increased effective diameter of the pore (ii) an
increased amount of open, functional pores or (iii) the absence of
sterical hindrance by the sugar and the loss of negatively charged
phosphate groups leading to reduced charge repulsion. From our data it
is obvious that the diameter of the pore does not change (Table
I), the number of open pores does not increase (Fig. 5), and therefore,
the absence of a sterical hindrance and reduced charge repulsion seem
to be more likely.
In Gram-negative bacteria, the periplasmic space is highly anionic
compared with the external medium, mainly because of the anionic
membrane-derived oligosaccharides. Membrane-derived oligosaccharides contribute to the Donnan potential across the outer membrane
VOM, which was determined for E. coli in
the presence of an external cation concentration of 100 mM
to be 26 mV (inside negative) (18). For the IM, a voltage of ~150 mV
(inside negative) is discussed (19).
For all lipid matrices used in this study, the voltage required for
voltage-dependent closure is |V| > 50 mV (Fig. 6,
A and B). Thus, from our data it may be concluded
that all PhoE porins incorporated into the OM are in the open state,
and therefore, they have no function for the regulation of the
transmembrane voltage. Interestingly, in the case of PL/PL membranes
the voltage required for closure of PhoE is 125 mV in both
directions, which is in the same order of magnitude as the proposed
transmembrane voltage across the IM. Similar values have also been
found in patch-clamp experiments on giant vesicles made from azolectin (33). Therefore, as was previously suggested (34), if trimers were
missorted to the IM, pores would be closed to prevent a short circuit
destroying the electrochemical potential across the IM. The IM is thus
protected by the following two mechanisms: (i) the reduced pore
formation in phospholipid membranes as compared with PL/LPS membranes
and (ii) the voltage-dependent closure of PhoE accidentally
incorporated into the IM. From the asymmetry of the
current/voltage characteristics of PhoE incorporated into symmetric
PL/PL membranes, a directed incorporation of the porin can be deduced
as it has been discussed in more detail with the porin from P. denitrificans (15), very likely with extracellular loops facing outwards.
For porin from P. denitrificans, a correlation between the
surface charge density of the glycolipid leaflet and porin gating was
found when the LPS side of the membrane was at a negative potential
with regard to the PL side (15, 16). These studies were, however,
restricted to Re LPS and GSL-1. In Fig. 6B it can be seen
that under the same conditions (positive voltage,
i.e. LPS side negative) PhoE channel closing is most
pronounced for LPS F515, which carries 4 negative charges, and
decreases over LPS R595 to LPS R45, which carries 3 negative charges.
However, for PhoE incorporated into the PL/LPS bilayers containing LPS from various strains of S. minnesota (Fig.
6A), channel closing is more efficient for PL/LPS R5
membranes followed by PL/LPS R7, and is least efficient in PL/LPS R595
membranes, given that all LPS carry 3 to 4 negative charges and the
increase in negative charge from these LPS to LPS R345 does lead to a
decreased gating compared with LPS R5. Furthermore, the gating was most
efficient for PL/PG membranes despite the fact that the absolute value
of the surface charge density is lower than for LPS R45.
Interestingly, in the positive voltage range the clamp voltage
necessary to induce pore closure of PhoE in lipid bilayers containing
on one side LPS from S. minnesota strains R595, R7, or R5
seems to decrease with increasing length of the sugar moiety. These
data indicate that not only for the pore formation but also for the
gating, not the surface charge density, but particular functional
groups of the LPS are decisive. Specific interactions between LPS and
the extracellular loops could influence their conformational changes,
which according to studies using atomic force microscopy are involved
in the gating process (35).
For negative clamp voltages down to 200 mV for all membranes made
from LPS from S. minnesota, almost no
voltage-dependent closure can be seen. This observation is
in complete contrast to the data obtained for LPS F515, LPS R45, and PG
showing a pronounced closing in the negative voltage range (Fig.
6B). This effect might be explained by mechanical stress
caused by the additional fatty acid of LPS from S. minnesota
compared with that from E. coli. An influence of the acyl
chain region of the lipid matrix on porin gating has also been observed
for porin from P. denitrificans (15) where gating was
completely inhibited at temperatures below the phase transition
temperature of the LPS. Thus, the additional fatty acid may have a
similar effect on the porin channel as acyl chain rigidification.
Our data clearly emphasize the role of LPS in pore formation by and
function of PhoE. In particular, from the qualitatively different
incorporation rates it can be deduced that LPS is the parameter
governing the formation of porin channels in the reconstituted OM and
that no particular insertion sites are required and, thus, LPS has an
important function for the late steps in porin biogenesis. The LPS
chemical structure does, however, not only influence pore formation,
but also channel gating. Thus, it is of unique importance to consider
the asymmetric composition of the OM of Gram-negative bacteria with LPS
on one side and PL on the other side when aiming at the determination
of porin function in reconstitution systems.
| |
FOOTNOTES |
*
This work was supported by the Deutsche
Forschungsgemeinschaft, SFB 470, Project B5.The costs of publication of this
article were defrayed in part by the
payment of page charges. The article must therefore be hereby marked
"advertisement" in
accordance with 18 U.S.C. Section
1734 solely to indicate this fact.
¶
To whom correspondence should be addressed: Research
Center Borstel, Center for Medicine and Biosciences, Dept. of
Immunochemistry and Biochemical Microbiology, Parkallee 10 D-23845 Borstel, Germany. Tel.: 49-4537-188-291; Fax:
49-4537-188-632; E-mail: awiese@fz-borstel.de.
Published, JBC Papers in Press, June 28, 2002, DOI 10.1074/jbc.M201950200
| |
ABBREVIATIONS |
The abbreviations used are:
OM, outer membrane;
IM, inner membrane;
LPS, lipopolysaccharide;
GSL-1, monosaccharide-type
fraction of glycosphingolipid;
PL, phospholipid mixture;
PE, phosphatidylethanolamine;
PG, phosphatidylglycerol;
DPG, diphosphatidylglycerol;
Kdo, 3-deoxy-D-manno-oct-2-ulosonic acid.
| |
REFERENCES |
| 1.
|
Lugtenberg, B.,
and Van Alphen, L.
(1983)
Biochim. Biophys. Acta
737,
51-115[Medline]
[Order article via Infotrieve]
|
| 2.
|
Rietschel, E. Th.,
Kirikae, T.,
Schade, F. U.,
Mamat, U.,
Schmidt, G.,
Loppnow, H.,
Ulmer, A. J.,
Zähringer, U.,
Seydel, U., Di,
Padova, F.,
Schreier, M.,
and Brade, H.
(1994)
FASEB J.
8,
217-225[Abstract]
|
| 3.
|
Holst, O.
(1999)
in
Endotoxin in Health and Disease
(Brade, H.
, Opal, S. M.
, Vogel, S. N.
, and Morrison, D. C., eds)
, pp. 115-154, Marcel Dekker, Inc., New York
|
| 4.
|
Nikaido, H.,
and Vaara, M.
(1985)
Microbiol. Rev.
49,
1-32[Free Full Text]
|
| 5.
|
Overbeeke, N.,
and Lugtenberg, B.
(1980)
FEBS Lett.
112,
229-232[CrossRef][Medline]
[Order article via Infotrieve]
|
| 6.
|
Benz, R.,
and Bauer, K.
(1988)
Eur. J. Biochem.
176,
1-19[Medline]
[Order article via Infotrieve]
|
| 7.
|
Cowan, S. W.,
Schirmer, T.,
Rummel, G.,
Steiert, M.,
Ghosh, R.,
Pauptit, R. A.,
Jansonius, J. N.,
and Rosenbusch, J. P.
(1992)
Nature
358,
727-733[CrossRef][Medline]
[Order article via Infotrieve]
|
| 8.
|
Bernstein, H. D.
(2000)
Curr. Opin. Microbiol.
3,
203-209[CrossRef][Medline]
[Order article via Infotrieve]
|
| 9.
|
Schatz, P. J.,
and Beckwith, J.
(1990)
Annu. Rev. Genet.
24,
215-248[CrossRef][Medline]
[Order article via Infotrieve]
|
| 10.
|
Tommassen, J.,
and Lugtenberg, B.
(1981)
J. Bacteriol.
147,
118-123[Abstract/Free Full Text]
|
| 11.
|
Ames, G. F. L.,
Spudich, E. N.,
and Nikaido, H.
(1974)
J. Bacteriol.
117,
406-416[Abstract/Free Full Text]
|
| 12.
|
De Cock, H.,
van Blokland, S.,
and Tommassen, J.
(1996)
J. Biol. Chem.
271,
12885-12890[Abstract/Free Full Text]
|
| 13.
|
De Cock, H.,
Brandenburg, K.,
Wiese, A.,
Holst, O.,
and Seydel, U.
(1999)
J. Biol. Chem.
274,
5114-5119[Abstract/Free Full Text]
|
| 14.
|
Sen, K.,
and Nikaido, H.
(1991)
J. Bacteriol.
173,
926-928[Abstract/Free Full Text]
|
| 15.
|
Wiese, A.,
Schröder, G.,
Brandenburg, K.,
Hirsch, A.,
Welte, W.,
and Seydel, U.
(1994)
Biochim. Biophys. Acta
1190,
231-242[Medline]
[Order article via Infotrieve]
|
| 16.
|
Wiese, A.,
Reiners, J. O.,
Brandenburg, K.,
Kawahara, K.,
Zähringer, U.,
and Seydel, U.
(1996)
Biophys. J.
70,
321-329[Medline]
[Order article via Infotrieve]
|
| 17.
|
Kawahara, K.,
Matsuura, M.,
and Danbara, H.
(1990)
Jpn. J. Med. Sci. Biol.
43,
250[Medline]
[Order article via Infotrieve]
|
| 18.
|
Sen, K.,
Hellman, J.,
and Nikaido, H.
(1988)
J. Biol. Chem.
263,
1182-1187[Abstract/Free Full Text]
|
| 19.
|
Bakker, E. P.,
and Mangerich, W. E.
(1981)
J. Bacteriol.
147,
820-826[Abstract/Free Full Text]
|
| 20.
|
Galanos, C.,
Lüderitz, O.,
and Westphal, O.
(1969)
Eur. J. Biochem.
9,
245-249[Medline]
[Order article via Infotrieve]
|
| 21.
|
Jansen, C.,
Heutink, M.,
Tommassen, J.,
and De Cock, H.
(2000)
Eur. J. Biochem.
267,
3792-3800[Medline]
[Order article via Infotrieve]
|
| 22.
|
Montal, M.,
and Mueller, P.
(1972)
Proc. Natl. Acad. Sci. U. S. A.
69,
3561-3566[Abstract/Free Full Text]
|
| 23.
|
Wiese, A.,
and Seydel, U.
(2000)
Methods Mol. Biol.
145,
355-370[Medline]
[Order article via Infotrieve]
|
| 24.
|
Osborn, M. J.,
Gander, J. E.,
Parisi, E.,
and Carson, J.
(1972)
J. Biol. Chem.
247,
3962-3972[Abstract/Free Full Text]
|
| 25.
|
Sen, K.,
and Nikaido, H.
(1991)
J. Biol. Chem.
266,
11295-11300[Abstract/Free Full Text]
|
| 26.
|
De Cock, H.,
and Tommassen, J.
(1996)
EMBO J.
15,
5567-5573[Medline]
[Order article via Infotrieve]
|
| 27.
|
Wiese, A.,
Grünewald, P.,
Schaper, K. J.,
and Seydel, U.
(2001)
J. Endotoxin Res.
7,
147-155[CrossRef][Medline]
[Order article via Infotrieve]
|
| 28.
|
Brandenburg, K.,
Kusumoto, S.,
and Seydel, U.
(1997)
Biochim. Biophys. Acta
1329,
183-201[Medline]
[Order article via Infotrieve]
|
| 29.
|
Weiss, M. S.,
Abele, U.,
Weckesser, J.,
Welte, W.,
Schiltz, E.,
and Schulz, G. E.
(1991)
Science
254,
1627-1630[Abstract/Free Full Text]
|
| 30.
|
Weiss, M. S.,
and Schulz, G. E.
(1992)
J. Mol. Biol.
227,
493-509[CrossRef][Medline]
[Order article via Infotrieve]
|
| 31.
|
Brandenburg, K.,
Richter, W.,
Koch, M. H. J.,
Meyer, H. W.,
and Seydel, U.
(1998)
Chem. Phys. Lipids
91,
53-69[CrossRef][Medline]
[Order article via Infotrieve]
|
| 32.
|
Korteland, J.,
and Lugtenberg, B.
(1984)
Biochim. Biophys. Acta
774,
119-126[Medline]
[Order article via Infotrieve]
|
| 33.
|
Berrier, C.,
Besnard, M.,
and Ghazi, A.
(1997)
J. Membr. Biol.
156,
105-115[CrossRef][Medline]
[Order article via Infotrieve]
|
| 34.
|
Eppens, E. F.,
Saint, N.,
Van Gelder, P.,
van Boxtel, R.,
and Tommassen, J.
(1997)
FEBS Lett.
415,
317-320[CrossRef][Medline]
[Order article via Infotrieve]
|
| 35.
|
Müller, D. J.,
and Engel, A.
(1999)
J. Mol. Biol.
285,
1347-1351[CrossRef][Medline]
[Order article via Infotrieve]
|
Copyright © 2002 by The American Society for Biochemistry and Molecular Biology, Inc.
 CiteULike  Complore  Connotea  Del.icio.us  Digg  Reddit  Technorati What's this?
This article has been cited by other articles:
|
|
|
|
 
E. Frirdich, C. Bouwman, E. Vinogradov, and C. Whitfield
The Role of Galacturonic Acid in Outer Membrane Stability in Klebsiella pneumoniae
J. Biol. Chem.,
July 29, 2005;
280(30):
27604 - 27612.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
E. Frirdich and C. Whitfield
Review: Lipopolysaccharide inner core oligosaccharide structure and outer membrane stability in human pathogens belonging to the Enterobacteriaceae
Innate Immunity,
June 1, 2005;
11(3):
133 - 144.
[Abstract]
[PDF]
|
|
|
Copyright © 2002 by the American Society for Biochemistry and Molecular Biology.
|
Advertisement
Advertisement
|
TOP
ABSTRACT
INTRODUCTION