(Received for publication, November 18, 1996, and in revised form, February 5, 1997)
From the Fachbereich Biologie/Chemie, Abteilung Mikrobiologie, Universität Osnabrück, D-49069 Osnabrück, Federal Republic of Germany
In response to K+ availability or medium osmolality, the sensor kinase KdpD and the response regulator KdpE control the expression of the kdpFABC operon, coding for the high affinity K+-translocating Kdp ATPase of Escherichia coli. The stimulus for KdpD to undergo autophosphorylation is believed to be a change in turgor or some effect thereof, reflecting the role of K+ as an important cytoplasmic osmotic solute. The membrane-bound sensor kinase KdpD was overproduced as a fusion protein containing six contiguous histidine residues two amino acids before the C terminus. This KdpD-His6 protein was functional in vitro and in vivo. KdpD-His6 was purified from everted membrane vesicles by solubilization with the zwitterionic detergent lauryldimethylamine oxide followed by nickel chelate chromatography and ion exchange chromatography to >99% homogeneity. The solubilized protein was not active with respect to autophosphorylation, but retained the ability to bind 2-azido-ATP. KdpD-His6 was reconstituted into proteoliposomes in a unidirectional inside-out orientation as revealed by ATP accessibility and protease susceptibility. Purified and reconstituted KdpD-His6 exhibited autokinase activity, and the phosphoryl group could be transferred to KdpE. Furthermore, KdpD-His6 was found to be the only protein that mediates dephosphorylation of KdpE~P.
Kdp is one of several transport systems that accumulate
K+ in bacteria. Among these systems, Kdp has the highest
affinity for K+, and it is only expressed when other
systems are unable to meet the cells' need for K+.
Expression of the kdpFABC operon, coding for the
K+-translocating Kdp ATPase, is regulated by two proteins
(KdpD and KdpE) that belong to the class of sensor kinase/response
regulator systems (see Refs. 1 and 2 for review). The kdpD
gene has been cloned, sequenced, and overexpressed (3). The gene
product KdpD is a membrane-bound protein located in the cytoplasmic
membrane (3, 4). According to hydropathy analysis of the primary amino
acid sequence, KdpD consists of an extended cytoplasmic N-terminal
region, four putative transmembrane domains, and an extended
cytoplasmic C-terminal region (see Fig. 1). This model is strongly
supported by the analysis of a series of KdpD-alkaline phosphatase
(kdpD-phoA) and KdpD--galactosidase
(kdpD-lacZ) fusions as well as protease susceptibility
studies (5).
Phosphorylation of KdpD is readily demonstrated in vitro by crude cell extracts (4, 6) and by partially purified protein after reconstitution (7). Phosphorylation proceeds most readily in the presence of Mg2+ and at a lower rate in the presence of Ca2+ (6). There is moderate homology of parts of the C-terminal domain to other sensor kinases, including His-673. When His-673 was replaced by glutamine, phosphorylation was no longer observed. This finding supports the view that the site of phosphorylation is His-673 (4). Addition of the soluble response regulator KdpE to phosphorylated KdpD results in rapid transfer of the phosphoryl group to KdpE (4, 6, 7). Furthermore, purified KdpE~P can be dephosphorylated in the presence of everted membrane vesicles containing KdpD (8). Analysis of a set of KdpD proteins carrying different truncations in the N-terminal part of the protein indicated that in vitro autophosphorylation is dependent on the membrane-spanning domains and that in vivo signal transduction is modulated by the N-terminal region of KdpD (8).
KdpD presumably functions as a turgor sensor. The turgor control model is based on the pattern of expression of kdpFABC in response to changes in medium K+ concentration and osmolality (9, 10), reflecting the role of K+ as an important cytoplasmic osmotic solute (11). However, further observations lend support to the notion that turgor is apparently not the sole regulatory signal for the expression of kdpFABC and that the availability or intracellular concentration of K+ may also be involved (12-14).
To address the question of the stimulus for KdpD under in vitro conditions and to characterize this protein biochemically, it is necessary to develop an efficient and rapid purification and reconstitution method for KdpD. This report describes the purification of KdpD to >99% homogeneity and an efficient reconstitution method. Purified KdpD in proteoliposomes was tested for its autophosphorylation, phosphotransfer, and dephosphorylation activities.
[-32P]ATP was purchased from
Amersham Buchler. All restriction enzymes and T4 DNA ligase were from
New England Biolabs Inc. Taq DNA polymerase was from Life
Technologies, Inc. Synthetic oligonucleotide primers were obtained from
Eurogentec. Alkaline phosphatase-conjugated goat anti-rabbit IgG was
purchased from BIOMOL Research Laboratories Inc.
Ni2+-NTA 1 resin was from
QIAGEN Inc. DEAE-Sepharose CL-6B and HiTrap desalting columns were from
Pharmacia Biotech Inc. Purified Escherichia coli lipids were
purchased from Avanti Polar Lipids. 2-Azido-[
-32P]ATP
was a gift from Jan Berden (University of Amsterdam). Centrifugal filter units were purchased from Millipore Corp. Bio-Beads SM-2 were
from Bio-Rad. Detergents were obtained from Calbiochem and Anatrace.
All other materials were reagent grade and obtained from commercial
sources.
E. coli JM109
(recA1 endA1 gyrA96 thi hsdR17 supE44
relA1
(lac-proAB)/F
traD36 proA+B+
lacIqlacZ
M15) (15) was used as carrier for the
plasmids described. E. coli TKR2000 (
kdpFABCDE thi
rha lacZ nagA trkA405 trkD1 atp706) (16) was used for expression
of kdpD from the tac promoter. E. coli
HAK006 (
kdpABCD
(lac-pro) ara
thi) carries a kdpFABC promoter-lacZ fusion
gene and is kdpE+ (17). This strain was used for
determination of in vivo signal transduction. Plasmid pPV5
was used in which kdpD is under the control of the
tac promoter (3).
To facilitate mutant
construction, a unique restriction site (HindIII) after the
stop codon of kdpD was created by polymerase chain reaction
using plasmid pPV5. Six codons for His were introduced at position 2676 of the kdpD gene by polymerase chain reaction using the
mutagenic antisense primer
5-AA-TCA-GAA-GCT-TTG-TCA-CAT-ATC-ATG-ATG-ATG-ATG-ATG-ATG-CTC-ATG-AAA-TTC-3
and a sense primer upstream of the ClaI restriction site.
The polymerase chain reaction product was purified on agarose gel, digested with ClaI and HindIII, and ligated to
similarly treated pPV5, resulting in pPV5-His6.
Mutations were verified by sequencing the length of the polymerase chain reaction-generated segment through the ligation junctions in double-stranded plasmid DNA using the dideoxynucleotide termination method (18) and synthetic sequencing primers after alkaline denaturation (19).
Preparation of Everted Membrane VesiclesE. coli
TKR2000 cells transformed with plasmid pPV5-His6 were grown
aerobically at 37 °C in KML complex medium (1% Tryptone, 0.5%
yeast extract, and 1% KCl) (20) supplemented with ampicillin (100 µg/ml). Cells were harvested at an absorbance at 600 nm of ~1.0.
Everted membrane vesicles were prepared by passage of cells through a
Ribi press and washed twice in an EDTA-containing buffer of low ionic
strength (21), except that the buffer was changed from Hepes/Tris to
Tris/HCl. The washed membrane vesicles were homogenized in 50 mM Tris/HCl, pH 7.5, 10% (v/v) glycerol, and 2 mM dithiothreitol; adjusted to 10 mg/ml protein; and stored until use at 80 °C.
To an aliquot
(typically 8 ml) of the washed membrane vesicles (5 mg/ml) was added
0.6 M NaCl, and while stirring on ice, KdpD-His6 was extracted with lauryldimethylamine oxide
(LDAO) that was added stepwise until a final concentration of 2% (v/v) was reached. After stirring on ice for 30 min, this solubilization mixture was centrifuged for 45 min at 264,000 × g in a
Beckman TL100 centrifuge. The supernatant fraction containing
solubilized KdpD-His6 was purified by affinity
chromatography on a 2-ml column of Ni2+-NTA resin in the
following manner. The resin was pre-equilibrated with 25 bed volumes of
column buffer (50 mM Tris/HCl, pH 7.5, 10% glycerol, 10 mM -mercaptoethanol, 0.5 M NaCl, 10 mM imidazole, and 0.04% (w/v) n-dodecyl
-D-maltoside (DM)). Binding of KdpD-His6 was
carried out batchwise by incubation of the solubilized proteins and the
Ni2+-NTA resin for 30 min at 4 °C. The protein-resin
complex was then packed into the column, and unbound protein was
removed by washing with column buffer until the absorbance at 280 nm
returned to base line. Bound KdpD-His6 was eluted by
increasing the imidazole concentration to 100 mM. The flow
rate was 1 ml/min, and 2-ml fractions were collected. KdpD-containing
fractions were pooled and desalted by passage through a HiTrap
desalting column. Final purification of KdpD-His6 was
achieved by loading the desalted protein on a 4-ml DEAE-Sepharose CL-6B
column pre-equilibrated with 50 mM Tris/HCl, pH 7.5, 10%
glycerol, 10 mM
-mercaptoethanol, and 0.04% (w/v) DM.
Contaminating proteins were removed by increasing the NaCl
concentration stepwise (0, 50, 100, 150, 200, and 300 mM).
KdpD-His6 was eluted at 600 mM NaCl.
KdpD-His6-containing fractions were pooled and used for
reconstitution.
Purified KdpD-His6 was reconstituted into E. coli phospholipids essentially as described (22). Liposomes were solubilized with the indicated detergents, preferentially DM. The process of liposome solubilization was analyzed by turbidity measurements. In a typical experiment, E. coli phospholipids (5 mg/ml) were solubilized with DM (0.58%, w/v). Then KdpD-His6 in elution buffer was added. The mixture was stirred at room temperature for 10 min. The final ratio of phospholipids to protein was kept at 25 (w/w). Bio-Beads were used to remove the detergent. Bio-Beads were thoroughly rinsed with methanol and buffer (23) and stored in 50 mM Tris/HCl, pH 7.5, and 10% glycerol at 4 °C until use. To remove the detergents, Bio-Beads at a bead/detergent ratio of 5 (w/w) were added, and the mixture was kept under gentle stirring at 4 °C. After incubation overnight, additional Bio-Beads were added to remove residual detergent. After further incubation for 3 h, the Bio-Beads were pipetted off. The turbid suspension was centrifuged for 1 h at 372,000 × g. The pellet was resuspended in 50 mM Tris/HCl, pH 7.5, 10% glycerol, and 2 mM dithiothreitol. Proteoliposomes were either used instantly or stored in liquid nitrogen.
Phosphorylation and Dephosphorylation AssaysEverted
membrane vesicles containing KdpD (2 mg/ml protein), solubilized KdpD
(2 mg/ml), or proteoliposomes (0.13 mg/ml) were incubated in
phosphorylation buffer containing 50 mM Tris/HCl, pH 7.5, 10% glycerol, 0.5 M NaCl, 10 mM
MgCl2, and 2 mM dithiothreitol at room
temperature. Phosphorylation was initiated by the addition of 20 µM [-32P]ATP (2.38 Ci/mmol) or 2 µM [
-32P]ATP (23.8 Ci/mmol) as
indicated. At different times, aliquots were removed and mixed with an
equal volume of twice concentrated SDS sample buffer (24). After
incubation for 5 min, an equimolar amount of purified KdpE (16.5 µg/ml) (8) was added to the KdpD-containing fractions, and the
incubation was continued. Additional aliquots were removed at different
times and mixed with SDS sample buffer as described above.
To test dephosphorylation, purified KdpE was phosphorylated in the
following manner. Wild-type KdpD (0.67 mg/ml) was incubated in
phosphorylation buffer containing 5 mM CaCl2
instead of MgCl2, and phosphorylation was initiated with 2 µM [-32P]ATP (23.8 Ci/mmol). After 10 min, purified KdpE (16.5 µg/ml) was added, and the incubation was
continued for 5 min. KdpD-containing membrane vesicles were removed by
two filtration steps. First, a membrane with a pore size of 0.22 µm
was used, followed by a second filtration through a low protein-binding
membrane that cuts off proteins with a Mr of
>30,000. ATP was removed by the addition of 20 mM
MgCl2, 3.2 mM glucose, and 5.4 units of
hexokinase. Dephosphorylation was initiated by the addition of
KdpD-His6 in everted membrane vesicles (0.44 mg/ml) or
proteoliposomes (0.09 mg/ml). At the times indicated, aliquots were
removed, and the reaction was stopped by the addition of SDS sample
buffer.
All samples were immediately subjected to SDS-polyacrylamide gel electrophoresis (PAGE) (24). Gels were dried, and phosphorylation of the proteins was detected by autoradiography with Eastman Kodak BIOMAX MR film for 14 h at room temperature.
Binding of 2-Azido-ATPBinding of
2-azido-[-32P]ATP was studied with everted membrane
vesicles containing KdpD-His6 (1 mg/ml) or solubilized
KdpD-His6 (1 mg/ml) in different detergents. The samples
were preincubated in 50 mM Tris/HCl, pH 7.5, 10% glycerol,
and 2 mM dithiothreitol. The reaction was initiated by the
addition of 52 µM 2-azido-[
-32P]ATP
(0.34 Ci/mmol) and incubated for 3 min. Cross-linking was achieved by
irradiation of the sample with UV light at 302 nm for 2.5 min. The same
experiment was repeated except that the samples were preincubated with
an excess of ATP (10 mM) for 3 min before
2-azido-[
-32P]ATP was added. All samples were
immediately subjected to SDS-PAGE. Gels were dried, and phosphorylation
of the proteins was detected by autoradiography.
For Western blot analysis, everted membrane vesicles or proteoliposomes were solubilized in an equal volume of twice concentrated SDS sample buffer. Equal amounts of protein were subjected to SDS-PAGE. Proteins were electroblotted, and immunoblots were probed with polyclonal antibodies against KdpD as described (4). Immunodetection was performed using an alkaline phosphatase-conjugated goat anti-rabbit IgG antibody (4).
Proteolysis with KallikreinProteoliposomes (0.2 mg/ml) in the corresponding buffer were incubated with kallikrein (100 µg/ml) for 1 h at 37 °C. At the indicated times, the reaction was stopped by the protease inhibitor leupeptin (1 µM), and the protein was subjected to SDS-PAGE and silver-stained.
Probing Signal Transduction in VivoIn vivo
signal transduction was probed with E. coli HAK006
transformed with the plasmids described. Cells were grown in TY medium
(1% Tryptone and 0.5% yeast extract) (14) supplemented with 0.6 M NaCl and KCl as indicated. The basal concentration of
K+ in the medium was determined to be 6 mM by
flame photometry. Cells were grown to the mid-logarithmic growth phase
and harvested by centrifugation. The cell pellet was resuspended in 200 mM sodium phosphate buffer, pH 7.2. An aliquot of this
suspension was used to determine the absorbance at 600 nm, and another
aliquot was used for determination of -galactosidase activity after
the cells were permeabilized with toluene.
-Galactosidase activity
was determined from at least three different experiments and is given in Miller units calculated as described (25).
The turbidity of the phospholipid vesicle suspension as a function of detergent concentration was measured at an absorbance of 600 nm with a Shimadzu spectrophotometer. Liposomes were suspended at a concentration of 5 mg/ml. The detergent concentration was raised in increments, and absorbance was continuously recorded until a steady-state level was reached.
Analytical MethodsProtein was assayed by a modified version (26) of the method of Lowry et al. (45) with bovine serum albumin as a standard. Proteins were separated by SDS-PAGE (24) using 7.5, 9, 11, and 12.5% acrylamide gels. Silver staining was performed as described (27).
Oligonucleotide-directed, site-specific mutagenesis was used to insert six consecutive triplets encoding histidine two triplets before the stop codon of kdpD (Fig. 1). This insertion was verified by double-stranded DNA sequencing. The gene encoding the tagged sensor kinase was cloned in pPV5 and overexpressed under the control of the tac promoter in E. coli TKR2000. KdpD-His6 was detectable in everted membrane vesicles in amounts that are comparable to wild-type KdpD as judged from immunoblot analysis (data not shown).
Properties of KdpD-His6 in Vitro and in VivoKdpD-His6 in everted membrane vesicles was able to undergo autophosphorylation in the presence of ATP. After the addition of the purified response regulator KdpE, a rapid transfer of the phosphoryl group from phosphorylated KdpD-His6 to KdpE occurred (data not shown). In addition, KdpD-His6 caused dephosphorylation of KdpE~P in a time-dependent manner. Simultaneously with the decrease in the amount of KdpE~P, an increase in the amount of phosphorylated KdpD-His6 was observed (data not shown). All activities tested with KdpD-His6 were indistinguishable from those of wild-type KdpD.
Furthermore, the sensing and signal transduction capacity of
KdpD-His6 in vivo was tested. For this purpose,
E. coli HAK006 was used (14), which lacks the functional
kdpFABC operon as well as kdpD on the chromosome,
but contains the intact kdpE gene under the control of its
own promoter. This strain carries a kdpFABC promoter-lacZ fusion gene on the chromosome. Upon
transformation of this strain with a plasmid carrying the intact
kdpD gene, the inducible expression of the
kdpFABC operon can be probed by measuring -galactosidase
activity. It has been found previously that the extent of
kdpFABC expression is dependent on the K+
concentration and the osmolality of the medium (9). E. coli HAK006 cells were transformed with plasmid pPV5 or
pPV5-His6 and grown in a medium containing 0.6 M NaCl and only basal levels of K+. The
steady-state
-galactosidase activity of these cells was high (110 units), indicating that KdpD as well as KdpD-His6 were able
to respond to high osmolality. Moreover,
-galactosidase activity was
reduced to basal levels (0.8 units) when the cells were cultivated in a
medium of the same osmolality containing 40 mM
K+, which was sufficient to repress expression of
kdpFABC. Thus, a significant difference between the
activities of KdpD-His6 and wild-type KdpD was observed
neither under in vivo nor under in vitro
conditions.
Efficient solubilization of the membrane-bound histidine kinase is one of the prerequisites for purification. To select a detergent meeting this criterion, everted membrane vesicles were incubated in the presence of different detergents at final concentrations exceeding their critical micelle concentrations. After a 30-min incubation on ice, the reaction mixture was centrifuged, and aliquots of the supernatant and the resuspended membrane pellet were analyzed by immunoblotting (data not shown). The zwitterionic detergents LDAO, Zwittergent 3-12, and Zwittergent 3-14 were found to be the most efficient detergents in solubilizing KdpD-His6. Furthermore, solubilization of KdpD-His6 could be increased to nearly 100% when the solubilization was carried out in the presence of NaCl. Nonionic detergents had limited solubilization efficiency. The efficiency increased with the length of the carbon chain: octyl glucoside < decyl maltoside < DM. Maximal solubilization with DM was ~50%.
When KdpD-His6 solubilized in LDAO, DM, or Zwittergent 3-12 was tested for autophosphorylation activity, no phosphorylation of the protein was detectable (data not shown). In addition to standard conditions, phosphorylation was also tested in the presence of different NaCl concentrations (0, 50, and 150 mM). The ATP concentration was also varied. However, no significant phosphorylation of KdpD-His6 was observed under any of the conditions tested (data not shown).
Furthermore, the ability of solubilized KdpD-His6 to bind
2-azido-[-32P]ATP was analyzed. As shown in Fig.
2, KdpD-His6 solubilized in DM was able to
bind 2-azido-ATP. As a comparison, binding of 2-azido-ATP to
KdpD-His6 in everted membrane vesicles is shown. More
important, 2-azido-ATP binding could not be observed when KdpD-His6 was solubilized in LDAO or Zwittergent 3-12 (data
not shown).
Purification of KdpD-His6
Since solubilization of
KdpD-His6 from membranes was nearly 100% using the
detergent LDAO in the presence of NaCl, this detergent was used for the
purpose of purification. Taking into account that only DM-solubilized
KdpD-His6 was able to bind 2-azido-ATP, the detergent was
changed from LDAO to DM in the first column chromatography step. The
protocol used for the purification of KdpD-His6 comprised
two steps: purification by nickel chelate affinity chromatography and
ion exchange chromatography. Experiments were generally carried out
starting from 40 mg of protein of everted membrane vesicles. Progress
of the purification was judged by SDS-PAGE analysis (Fig.
3).
In a typical experiment, everted membrane vesicles were solubilized in the presence of 2% LDAO and 0.6 M NaCl as described under "Experimental Procedures." As judged by the protein content, solubilization with LDAO was not very selective. Thus, 90% of the membrane proteins were solubilized by this detergent. The highest degree of purification of KdpD-His6 was achieved by the nickel chelate chromatography. Binding of KdpD-His6 to the Ni2+-NTA resin was most effective when the solubilized proteins were incubated batchwise with pre-equilibrated resin for 30 min. Furthermore, selective binding of KdpD-His6 was greatly enhanced by the addition of low amounts of imidazole (10 mM) and relatively high amounts of NaCl (0.5 M). When the absorbance at 280 nm of the elute returned to base line, bound KdpD-His6 was eluted with 100 mM imidazole. As demonstrated by silver staining (Fig. 3), the fraction eluted with 100 mM imidazole shows a major protein band at ~98 kDa that corresponds to KdpD-His6 and some other minor bands.
To remove the residual contaminating proteins, ion exchange chromatography was used as a last purification step. Before loading the protein on a DEAE-Sepharose CL-6B column, the ionic strength of the pooled Ni2+-NTA column fractions was lowered by passage through a HiTrap desalting column. KdpD-His6 was bound to the ion exchanger. By applying a stepwise gradient of NaCl, contaminating proteins were eluted, whereas KdpD-His6 still remained bound to the resin. Finally, KdpD-His6 could be eluted at a NaCl concentration of 0.6 M. The eluted protein was subjected to SDS-PAGE, and after silver staining, only one band was detectable (Fig. 3). This protein band was identified as KdpD by Western blot analysis (Fig. 3). KdpD-His6 appeared as a single band; no dimer formation or even higher order aggregates were observed. Overall, starting from 10 g of cells (wet weight), 1.5 mg of purified KdpD-His6 was obtained.
Reconstitution of KdpD-His6Fractions containing
KdpD-His6 eluted with 0.6 M NaCl from the
DEAE-Sepharose column were pooled, and the protein was reconstituted into E. coli phospholipids. Reconstitution was carried out
using the detergent-mediated method as described (22). Liposomes were solubilized with increasing concentrations of DM, and the process of
liposome solubilization was followed by turbidity measurements. As
described for other membrane proteins, the stage of solubilization of
liposomes prior to the addition of solubilized protein results in
proteoliposomes of different activity (22). For DM, two characteristic solubilization stages of the liposomes were tested: partial
solubilization and total solubilization. Purified KdpD-His6
in elution buffer was added to the solubilized liposomes, and the
detergent was removed by the addition of Bio-Beads. Upon reconstitution
of KdpD-His6 into proteoliposomes, autophosphorylation
activity was recovered (Fig. 4A). The highest
phosphorylation activity was achieved with liposomes that were totally
solubilized with DM, as shown in Fig. 4A. Additional
detergents were tested for their ability to mediate reconstitution of
KdpD-His6. Thus, LDAO, Triton X-100, sodium cholate, and
octyl glucoside were used, and different stages of liposome
solubilization were tested. Although all prepared
KdpD-His6-containing proteoliposomes exhibited
autophosphorylation activity (data not shown), the highest activity was
always observed with liposomes that were solubilized with DM. The
efficiency of KdpD-His6 reconstitution was determined to be
35%.
Properties of Purified KdpD-His6 in Proteoliposomes
In addition to testing the autophosphorylation
activity of purified KdpD-His6 in proteoliposomes, transfer
of the phosphoryl group to purified KdpE was examined. Therefore,
proteoliposomes were incubated with [-32P]ATP, and
after 5 min, an equimolar amount of purified response regulator KdpE
was added. Subsequent transfer of the phosphoryl group to KdpE occurred
(Fig. 4A). To determine whether purified KdpD-His6 would regenerate KdpE from its phosphorylated
form, KdpE~P was incubated with KdpD-His6 in
proteoliposomes. A time-dependent dephosphorylation of
KdpE~P was observed (Fig. 4B). Dephosphorylation of
KdpE~P was accompanied by an increase in the amount of phosphorylated KdpD-His6 (Fig. 4B).
The
accessibility of ATP to KdpD-His6 in proteoliposomes
suggested already an inside-out orientation of this protein. To obtain additional evidence for this, the susceptibility to the protease kallikrein was tested. In both the N- and C-terminal regions of KdpD,
but not in the central hydrophobic region, 10 potential cleavage sites
for kallikrein can be found. Earlier studies revealed that treatment of
spheroplasts with kallikrein did not affect the kinase at all, but
after permeabilization of the spheroplasts, KdpD was degraded
completely, indicating that the protease cleavage sites are accessible
only from inside the cells (5). When proteoliposomes were incubated
with kallikrein, KdpD-His6 was completely degraded (Fig.
5). This result provides strong evidence for the
unidirectional inside-out orientation of KdpD-His6 in
proteoliposomes.
Most histidine kinase molecules in bacteria, including KdpD, are integral membrane proteins (28, 29). To study their activity in vitro, it is therefore necessary to establish an efficient solubilization and purification protocol. In the case of KdpD, this was achieved by applying the histidine peptide fusion technique (30). Although the His tag is located in the C-terminal domain, which is most certainly responsible for the interaction with ATP and the cytoplasmic response regulator KdpE, no significant changes with respect to phosphorylation, phosphotransfer, or dephosphorylation in comparison to the wild type could be observed. In addition, sensing and signal transduction in vivo were not influenced by the insertion of six His residues in KdpD as shown by the pattern of expression of the kdpFABC promoter-lacZ fusion gene in E. coli HAK006. To our knowledge, there is, so far, no other example of a fusion at the C-terminal end of a sensor kinase. In contrast, different sensor kinases with modifications at the N terminus have been described. Thus, the soluble parts of the sensor kinases NarX and NarQ were fused at the N-terminal end with six His residues, and the purified proteins were active (31). In NRII, the N-terminal domain was replaced by the maltose-binding protein. This protein still catalyzed autophosphorylation and transfer of the phosphoryl group to the response regulator (NRI), but it was unable to stimulate the dephosphorylation of NRI~P (32).
Solubilization of KdpD-His6 in different detergents
resulted in an inactive protein. This result is in agreement with
earlier attempts to detect autophosphorylation activity of the
solubilized protein (7). Obviously, only the membrane-bound protein
exhibits autophosphorylation activity. This assumption is further
supported by the analysis of a number of truncated KdpD proteins (8). Different truncations of the N-terminal part resulted in KdpD proteins
that retain autophosphorylation and phosphotransfer activities in
vitro. In contrast, phosphorylation could not be detected in those
forms of KdpD in which the membrane-spanning segments were removed. It
is important to note that these inactive truncated forms of KdpD could
be labeled with 2-azido-[-32P]ATP, indicating that
binding of ATP to KdpD was not affected by removal of the hydrophobic
domain. Our studies demonstrate that DM-solubilized
KdpD-His6 also retains its ability to bind 2-azido-ATP.
This observation revealed that binding of ATP is not dependent on the
anchoring of the protein in the membrane. In contrast, 2-azido-ATP
binding was not detectable when KdpD-His6 was solubilized
in zwitterionic detergents. Obviously, DM stabilizes the structure of
KdpD, an effect that was also observed for other membrane proteins
(e.g. Ref. 33).
Reconstitution of KdpD-His6 into proteoliposomes restores full activity of the sensor kinase. Therefore, KdpD-His6 in proteoliposomes was phosphorylated when incubated in the presence of ATP, and the phosphoryl group could be transferred to the soluble response regulator KdpE. That these activities of KdpD are dependent on a membranous environment is atypical for membrane-bound sensor kinases. Soluble truncated forms of the sensor kinases FixL (34), EnvZ (35, 36), and NarX and NarQ (31) are capable of undergoing autophosphorylation. The finding that anchoring of KdpD in the membrane is a prerequisite for an active conformation is in accordance with the current model for the regulation of kdpFABC expression (1). When turgor (and hence, membrane stretch) is high, KdpD is inactive. When turgor is reduced, KdpD alters its conformation, and the kinase becomes active, resulting in the formation of KdpE~P and thus expression of kdpFABC.
However, it is also conceivable that the key feature of regulation is the dephosphorylation of KdpE~P. Purified and reconstituted KdpD-His6 is able to facilitate dephosphorylation of purified KdpE~P. Thus, dephosphorylation of KdpE~P is dependent only on the histidine kinase KdpD. Dephosphorylation of the response regulator has been well studied in other sensor kinase/response regulator systems and characterized by diverse mechanisms (37). The sensor kinase EnvZ by itself is able to enhance the rate of dephosphorylation of phospho-OmpR (38, 39). In the case of NRII, an auxiliary protein (PII) is involved (32, 40, 41). In the chemotaxis system, dephosphorylation of CheY is mediated by a separate protein, CheZ (42), and in the process of sporulation of Bacillus subtilis, different phosphatases have been identified (43).
Dephosphorylation of KdpE~P is accompanied by an increase in the
amount of phosphorylated KdpD. Whether this is due to a reverse phosphotransfer of the phosphoryl group from KdpE to KdpD or to residual amounts of ATP is still under investigation. For the EnvZ/OmpR
system, a reverse phosphotransfer of the phosphoryl group from OmpR to
the autophosphorylation site of a
kinase/phosphatase+ mutant of EnvZ was
recently reported (44).
In summary, we have shown that the purified and reconstituted KdpD-His6 protein retains enzymatic activity. KdpD has an autokinase activity, and the phosphoryl group can be transferred to the response regulator KdpE. Finally, KdpD facilitates the dephosphorylation of purified KdpE~P. This study provides the basis for further analyses of the structure and function of KdpD. In this context, the question of how KdpD senses changes in turgor is especially interesting.
We thank Dr. T. Mizuno for providing E. coli strain HAK006 and M. Lucassen for providing purified response regulator KdpE.