(Received for publication, February 26, 1997)
From the Forschungsstelle "Enzymologie der
Proteinfaltung" der Max-Planck-Gesellschaft zur Förderung der
Wissenschaften e.V., Kurt-Mothes-Str. 3, 06120 Halle/Saale, Germany,
the § Martin-Luther-Universität Halle-Wittenberg,
Institut für Geologische Wissenschaften und Geiseltalmuseum,
Domstr. 5, 06108 Halle/Saale, Germany, and the
¶ Universität Zürich, Biochemisches Institut,
Winterthurerstr. 190, 8057 Zürich, Switzerland
In Escherichia coli as many as nine
different genes coding for proteins with significant homology to
peptidyl-prolyl cis/trans-isomerases (PPIases) have been
found. However, for three of them, the histidine-rich SlyD, the
homologous gene product of ORF149, and parvulin-like SurA, it was not
known whether these proteins really possess PPIase activity. To gain
access to the full set of PPIases in E. coli, SlyD, the
N-terminal fragment of SlyD devoid of the histidine-rich region, as
well as the protein product of ORF149 of E. coli named SlpA
(SlyD-like protein) were cloned,
overexpressed, and purified to apparent homogeneity. On the basis of
the amino acid sequences, both proteins proved to be of the
FK506-binding protein type of PPIases. Only when using trypsin instead
of chymotrypsin as helper enzyme in the PPIase assay, the enzymatic
activity of full-length SlyD and its N-terminal fragment can be
measured. For Suc-Ala-Phe-Pro-Arg-4-nitroanilide as substrate,
kcat/Km of 29,600 M1 s
1 for SlyD and 18,600 M
1 s
1 for the N-terminal
fragment were obtained. Surprisingly, the PPIase activity of SlyD is
reversibly regulated by binding of three Ni2+ ions to the
histidine-rich, C-terminal region. Because the PPIase activity of SlpA
could be established as well, we now know eight distinct PPIases with
proven enzyme activity in E. coli.
Peptidyl-prolyl cis/trans-isomerases (PPIases,1 EC 5.2.1.8.) are able to accelerate the cis-trans-isomerization of the Xaa-Pro peptide bond in polypeptide chains (1). They are divided into three families, the cyclophilins, FK506-binding proteins (FKBP) including the putative subfamily of trigger factors, and parvulins. Members of all three PPIase families have been found in Escherichia coli (2-6).
The slyD gene (sensitivity to lysis)
was identified by screening for survival of E. coli C after
induction of the cloned lysis gene E of bacteriophage
X174 (7). Mutations mapped to the slyD locus (at 73.5
on
the E. coli chromosome) resulted in accumulation of complete
phages inside the cells, but no lysis occurred. The slyD
locus was cloned and sequenced. By deletion analysis it was demonstrated that the open reading frame 196 of E. coli is
identical to slyD (8). The derived amino acid sequence of
SlyD was shown to share similarity with FKBPs. Independently, the
protein SlyD (there called WHP, wonderous
histidine-rich protein) was discovered by binding
to nickel ions immobilized on nitriloacetic acid-agarose (NTA) resin
(9).
Derived from the amino acid sequence, SlyD consists of two sequence regions, an N-terminal stretch of 146 amino acids with 28.1% similarity to hFKBP12 and a C-terminal histidine-rich stretch. Similar regions containing clustered histidine residues are usually present in HypB proteins involved in nickel insertion into hydrogenases (10). Only HypB of E. coli lacks this histidine-rich stretch. In contrast to HypB of other species, it is not able to bind nickel ions (11). A hypothetical protein with a similar structure as SlyD was detected in Haemophilus influenzae (12).
The deduced amino acid sequence of the ORF149 from E. coli
(13) is homologous to amino acids 1-146 of SlyD. A protein, perhaps derived from ORF149 of E. coli, was found earlier after
cloning of a 7.1 kilobase insert generated by restriction of the
specialized transducing phage dapB2 with the restriction
endonuclease BamHI (14), and a 17-kDa protein was detected
by coupled transcription and translation. After sequencing of the
lsp-dapB interval (13), it was suggested that the 17-kDa
protein is coded by ORF149. The ORF149 is part of the
ileS-lsp operon (15). An operon of the same structure,
x-ileS-lsp-orf149-orf316, was found in Enterobacter aerogenes (16). A similar ORF has also been reported for
Pseudomonas fluorescens (17) and for Methanococcus
jannaschii (18).
Here we describe that SlyD is indeed a PPIase. Its enzymatic activity can be reversibly inhibited by Ni2+ ions. Metal ion influence disappears for the N-terminal PPIase fragment of SlyD (amino acids 1-146) lacking the histidine-rich region. Just like SlyD, the related protein product of ORF149 of E. coli (SlpA) shows PPIase activity.
All genes or part of the genes were amplified from E. coli strain BL21 chromosomal DNA by polymerase chain reaction (PrimeZyme, Biometra, Göttingen, Germany) and were ligated into the overexpression plasmid pQE30 (Qiagen, Hilden, Germany) followed by transformation into competent E. coli M15(pREP4) cells. The cells were selected on Luria-Bertani agar plates containing 50 µg/ml ampicillin and 10 µg/ml kanamycin. The plasmids of the obtained clones were examined by restriction analysis and DNA sequencing. In addition, the clones were screened for overexpression of a protein of the right size in small scale cultures. Three h after induction by 1 mM IPTG, the cells were pelleted by short centrifugation, dissolved in sample buffer, and analyzed by SDS-PAGE (19) and Coomassie Blue staining (20).
For cloning of the slyD gene, the oligonucleotide
primers 5-CGCCAGAATTCTTCCCATGCTCAGG-3
and
5
-GCGTCTGGGATCCGAGAAGTTAAAAGCCAGCC-3
were used. The obtained
fragment was cloned into pQE30 by using the restriction endonucleases
EcoRI and BamHI. For cloning of the N-terminal
fragment of SlyD (amino acids 1-146), the same cloning strategy as for
SlyD with the above mentioned first primer and the primer
5
-GTGAAGGATCCTCAGCTAATTATTCTTCTTCAGTCGC-3
was used. Cloning of the
ORF149 was achieved by polymerase chain reaction using the primers
5
-CGGCAAAGAATTCTAAATATAAG AGC-3
and
5
-CGCGAGCTCTCGACCACATAGCG-3
followed by ligation of the fragment
into the EcoRI and SacI restriction sites of
pQE30. All DNA manipulations were performed according to Sambrook
et al. (21). DNA sequencing of recombinant plasmids was performed by the Sanger method of chain termination using the T7
sequencing kit (Pharmacia, Freiburg, Germany) and the pQE type III/IV
or reverse primers (Qiagen, Hilden, Germany). All DNA modifying
enzymes were purchased from Boehringer Mannheim (Germany).
Overexpression of recombinant proteins was achieved by inoculating 1 liter of medium containing 50 µg/ml ampicillin and 10 µg/ml kanamycin with 20 ml of overnight culture of the appropriate strain and growing at 37 °C and 200 rpm shaking to an A600 of 0.5. After induction with 1 mM IPTG, the cells were grown for another 5 h. The cells were harvested at 4,000 × g for 10 min in a Beckman J2-HC centrifuge (Beckman Instruments, Palo Alto, CA) and dissolved in the buffer used for purification on the first column. For degradation of nucleic acids, 0.1% (v/v) benzonase was added. The cell suspension was passed 3 times through a French press cell (SLM Aminco, Buettelborn, Germany) at 20,000 p.s.i. After the removal of membrane debris by centrifugation at 95,800 × g for 40 min at 4 °C in a Beckman L8 60 M centrifuge, the supernatant was submitted to further purification.
Determination of Protein Concentration, SDS-PAGE, Western BlottingFor determination of the protein concentration, the
method of Bradford (22) and the absorbance at 280 nm were used. The extinction coefficients at 280 nm were calculated according to Gill
and van Hippel (23). Polyacrylamide gel electrophoresis was performed
as described by Laemmli (19), and the protein bands were visualized by
silver staining (24) or a modified Coomassie Blue staining method (20).
Polyclonal antibodies against SlyD and SlpA were raised in rabbits (PAB
productions, Hebertshausen, Germany) against proteins desalted by HPLC
as described below. Western blotting to nitrocellulose was performed in
a semidry electroblotting system (Fast Blot B32, Biometra,
Göttingen, Germany) with a buffer containing 25 mM
Tris, 150 mM glycine, 10% (v/v) methanol, pH 8.3. The
membrane was incubated with diluted polyclonal antibodies (1:2,000) for
1 h at room temperature. For detection, the blot was incubated
with peroxidase-conjugated goat anti-rabbit IgG (Sigma, Deisenhofen,
Germany) and developed with 0.018% 4-chloro-1-naphthol in 50 mM Tris, 150 mM NaCl, 0.024% hydrogen
peroxide, pH 6.0.
SlyD was purified by affinity chromatography on Ni2+-NTA-agarose (Quiagen, Hilden, Germany) with a 0-100 mM imidazole gradient in 100 mM Tris, 1 M NaCl, pH 8.0. After dialysis against 10 mM MOPS, 1 mM dithiothreitol, pH 7.0, an anion exchange chromatography on Fractogel EMD DEAE-650(M) was performed. The SlyD containing fractions were saturated to 30% with ammonium sulfate and applied on a propyl-Sepharose column. The protein was eluted with a gradient of 30-0% saturation ammonium sulfate and dialyzed thoroughly against 10 mM MOPS, 1 mM dithiothreitol, pH 7.0. After that the protein was enriched on a DEAE column. About 1 mg of homogeneous SlyD was desalted by HPLC and used to generate antibodies in rabbit. All protein preparations were performed at 4 °C. Purification was examined by SDS-PAGE and Western blotting.
The N-terminal fragment of SlyD was purified by applying the bacterial lysate on a Fractogel EMD DEAE-650(M) column (Merck, Darmstadt, Germany) equilibrated with 50 mM Tris, pH 7.5. The protein eluted with a 0-2 M NaCl gradient and was passed in the same buffer through a column packed with chelating Sepharose Fast Flow (Pharmacia, Uppsala, Sweden). Beforehand, the column was loaded with NiSO4 and washed with 50 mM Tris, pH 7.5. The protein fraction, which did not bind to the column, was dialyzed in 20 mM MES, pH 5.5, and passed through a Fractogel TSK AF-Blue column in the same buffer. The protein solution was concentrated on a Filtron Omegacell (exclusion size 10,000 Da). Samples of 1 ml were submitted to gel filtration on a Superdex 75 16/60 HiLoad FPLC column (Pharmacia, Uppsala, Sweden) equilibrated with 10 mM HEPES, 150 mM KCl, pH 7.8. After dialysis against 20 mM MES, pH 5.5, the appropriate fractions were applied in dialysis buffer on an EMD DEAE-650(M) column. The protein was eluted with a linear gradient from 0-2 M NaCl.
Purification of SlpA was achieved by anion exchange chromatography on Fractogel EMD DEAE-650(M) with 20 mM Tris, pH 7.5. The protein eluted with a linear gradient from 0-2 M NaCl and was passed through a chelating Sepharose Fast Flow column loaded with NiSO4 and equilibrated with 20 mM HEPES, 300 mM NaCl, pH 7.0. After concentration on a Filtron Omegacell (exclusion size 10,000 Da) the protein was submitted to gel filtration on a Superdex 75 column, equilibrated with 10 mM HEPES, 150 mM KCl, pH 7.8. The protein was dialyzed against 20 mM MES, pH 5.5, and passed through a Fractogel TSK AF-Blue column.
Measurement of PPIase ActivityThe protease coupled assay of Fischer et al. (1, 25) for determination of PPIase activity was used. Before any measurements were carried out, the protease stability of the possible PPIases was examined. About 100 µg/ml homogenous protein was incubated for either 20 s or 2 min in 35 mM HEPES, pH 7.8, at 10 °C with chymotrypsin (30 µM), subtilisin (2 µM), thrombin (60 µM), and trypsin (0.5 µM). The reaction was stopped by adding the serine protease inhibitor phenylmethanesulfonyl fluoride (1 mM final concentration) and boiling sample buffer. The samples were analyzed by SDS-PAGE and Western blotting.
For the determination of PPIase activity either 0.5 µM trypsin or 30 µM chymotrypsin and substrates of the structure Suc-Ala-Xaa-Pro-Arg-4NA and Suc-Ala-Xaa-Pro-Phe-4NA (Xaa = variable aminoacyl residue) were used, respectively. Stock solutions of substrates (10 mg/ml) were prepared in dimethyl sulfoxide. The reactions were carried out in 35 mM HEPES, pH 7.8, at 10 °C and followed spectrophotometrically at 390 nm using a Hewlett-Packard 8452 diode array UV/VIS spectrophotometer. 500 data points were used to calculate the first-order rate constants. Inhibition of PPIase activity was measured by adding FK506, CsA, or metal chlorides to the reaction mixture 5 min before starting the measurement. Stock solutions of 10 mM FK506 (gift from Fujisawa Pharmaceutical Co., Osaka, Japan) or 100 mM CsA in ethanol, 1 mM NiCl2, or other metal chlorides (ZnCl2, CaCl2, NaCl, KCl, and MgCl2) in 35 mM HEPES, pH 7.8, were used.
Analysis of the Nickel Ion Binding to SlyDEquilibrium dialysis was performed in home-made cells (26) for 12 h in a humid chamber at 10 °C. The lower chamber contained 250 µl of the appropriate NiCl2 solution in 35 mM HEPES, pH 7.8. The upper chamber was filled with 250 µl of SlyD solution (72 µM) in the same buffer.
The nickel content of SlyD and buffer samples was determined by inductively coupled plasma AES with a Jenaquant AES 110 (Zeiss, Jena, Germany). For calibration, 1-element standards (Merck, Darmstadt, Germany) and 35 mM HEPES, pH 7.8, as metal-free control have been used.
The data obtained by AAS were analyzed by means of the Adair equation (27).
![]() |
(Eq. 1) |
The data from the inhibition studies of the PPIase activity of SlyD by NiCl2 were converted to binding data. The specific PPIase activity of SlyD samples without added NiCl2, as these prepared for AES, should equal 0.24 bound nickel ions/molecule of SlyD (as determined by AES for these samples). All other data points were calculated, assuming linear dependence of nickel binding to inhibition of PPIase activity of SlyD.
CD Spectrometry of SlyDHomogeneous samples of SlyD were applied to a desalting column (Econo-Pac® 10 DG, Bio-Rad, Hercules, USA). The protein was eluted with 5 mM phosphate buffer, pH 6.0. CD spectra were recorded from 190 to 260 nm with 12.5 µM SlyD. When required, 100 µM NiCl2 (final concentration) was added 5 min before starting the measurement. All measurements were performed in a 1-mm cuvette at 25 °C with a J-710 spectropolarimeter (Jasco, Tokyo, Japan), averaging 20 transients with Jasco software Version 1,33.00.
HPLC, Mass Spectrometry and Protein SequencingReversed-phase HPLC separations for antibody production, protein sequencing, and mass spectrometry were performed on a Shimadzu LC-10A modular HPLC unit (Shimadzu, Kyoto, Japan). The protein was applied to a C3 column (125 × 4 mm Nucleosil 500-5 C3-PPN, Macherey-Nagel, Düren, Germany) in 1% acetonitrile, 0.1% trifluoroacetic acid and eluted with a 30-60% acetonitrile gradient in 0.1% trifluoroacetic acid for 20 min at 40 °C (flow rate 1 ml/min). The molecular mass of proteins embedded in sinapic acid was determined using a MALDI-TOF Bruker reflex mass spectrometer (Bruker-Franzen-Analytik, Bremen, Germany) with an N2-UV laser (337 nm). Amino acid sequences were determined using an Applied Biosystems 476A gas phase sequencer (Applied Biosystems, Foster City, USA).
SlyD was purified from overexpression cultures
with a yield of 54 mg of homogenous protein/liter of culture. By silver
staining, only a single band could be detected in SDS-PAGE after the
last purification step. Amino acid sequencing of the purified protein for 10 steps showed the expected N-terminal sequence MKVAKDLVVS of SlyD
(Fig. 1). The N-terminal methionine was not processed. Analysis of the protein by mass spectrometry showed a molecular mass of
20,850 Da, very similar to the theoretical mass of 20,852 Da. The
apparent molecular mass in SDS-PAGE was 25 kDa.
The purification of the N-terminal fragment of SlyD yielded 9 mg of homogenous protein/liter of culture. The protein did not bind to nickel ions immobilized on chelating Sepharose. In SDS-PAGE, the apparent molecular mass of the N-terminal fragment of SlyD was 18 kDa. Antibodies raised against the complete SlyD molecule were able to detect the N-terminal fragment of SlyD in Western blots.
The ORF149 of E. coli was cloned, and the published sequence (13) was verified. The protein product (SlpA) was purified to homogeneity as shown by SDS-PAGE and Coomassie Blue staining at a level of 17 mg/liter of culture. SlpA passed through a chelating Sepharose column preloaded with nickel ions. The measured molecular mass of 15,946 Da was in good agreement with the calculated value of 15,949 Da for SlpA if the methionine in position 1 has been cleaved off during processing. With antibodies generated against recombinant SlpA, a protein of the same apparent molecular mass as SlpA was detected in E. coli BL21.
PPIase Activity of SlyD and Related ProteinsIn previous
experiments using the standard assay coupled with chymotrypsin (1, 25),
no PPIase activity of SlyD has been detected (9). We found that this is
due to a high sensitivity of SlyD against chymotrypsin. After 2 min of
incubation with 30 µM chymotrypsin or 2 µM
subtilisin, SlyD was no longer detected in the Western blot. Even after
20 s of incubation, the amount of SlyD was considerably reduced.
In contrast, SlyD remained stable for at least two min toward 0.5 µM trypsin and 60 µM thrombin. For that
reason, PPIase activity measurements were carried out with trypsin,
another isomer-specific protease. For the substrates Suc-Ala-Phe-Pro-Arg-4NA, Suc-Ala-Ala-Pro-Arg-4NA, and Suc-Ala-Leu-Pro- Arg-4NA, specificity constants
kcat/Km of 29,600, 6,200, and
5,600 M1 s
1 were measured.
After addition of chymotrypsin (final concentration 30 µM) to the reaction mixture, no PPIase activity was
detected for SlyD, explaining the failure to detect it in the standard assay.
As shown earlier (9), SlyD is able to bind Ni2+ and
Zn2+ ions. The PPIase activity of SlyD was considerably
reduced in a reversible manner by addition of NiCl2 at
micromolar concentrations (Fig. 2). Incubation of a
sample of SlyD (1 µM) with 50 µM
NiCl2 decreased the PPIase activity of SlyD to 10%. The
activity could be completely restored by dilution into nickel-free
buffer. Other metal chlorides tested (CaCl2, NaCl, KCl,
MgCl2) did not affect the PPIase activity up to 1 mM final concentration. However, it was not possible to examine the influence of ZnCl2 due to the inactivation of
trypsin in the protease coupled assay. The PPIase activity of SlyD was inhibited neither by 25 µM FK506 nor by 100 µM CsA, otherwise typical inhibitors of PPIases, in the
reaction mixture.
The PPIase activity of the N-terminal fragment of SlyD measured
using trypsin-mediated cleavage of the substrates
Suc-Ala-Phe-Pro-Arg-4NA, Suc-Ala-Ala-Pro-Arg-4NA, and
Suc-Ala-Leu-Pro-Arg-4NA was found to be 18,600, 3,900, and 3,600 M1 s
1, respectively. This
activity was also not inhibited by 25 µM FK506 or 100 µM CsA. No inhibition of PPIase activity was detected up
to 100 µM NiCl2. The N-terminal
fragment of SlyD showed the same sensitivity to chymotrypsin as the
complete protein.
The purified protein SlpA was tested for PPIase activity in the
chymotrypsin coupled assay. In contrast to SlyD, SlpA remained active
even after 2 min of incubation with chymotrypsin. The protease resistance was also confirmed in SDS-PAGE. For the substrate
Suc-Ala-Phe-Pro-Phe-4NA, a specificity constant
kcat/Km of 7,400 M1 s
1 was determined. The
substrate specificity in comparison with other PPIases from E. coli is summarized in Table I. The PPIase activity
of SlpA was also measured with trypsin and the substrate Suc-Ala-Phe-Pro-Arg-4NA and was found to be 6,000 M
1 s
1. FK506 up to 25 µM, CsA up to 100 µM, and NiCl2
up to 100 µM did not decrease the PPIase activity of
SlpA.
|
For interpretation of PPIase activity, the amino acid sequences of the N-terminal fragment of SlyD and SlpA were compared with the sequences of trigger factor and human FKBP12 (Fig. 1). The N-terminal fragment of SlyD shows only weak homology to both hFKBP12 (13.7% identity, 28.1% similarity) and trigger factor (15.1% and 27.4%, respectively). SlpA is 33.5% similar to hFKBP12 (18.1% identity) and 27.6% similar to trigger factor (12.8% identity).
Analysis of Nickel Binding to SlyDFrom AES measurements, 3 bound nickel ions/SlyD molecule were observed (Fig. 3).
The binding constants K1,
K2, and K3 for the three
different binding sites were calculated to be 9.5 × 105 M1, 4.9 × 105 M
1 and 4.4 × 105 M
1. The Kd
value was calculated to 1.8 µM NiCl2
(inflection point of the binding function).
CD spectra of SlyD were recorded without and with addition of 100 µM NiCl2. The molar ellipticity spectra in
the range from 190 to 260 nm are shown in Fig. 4. The
maximum at 200 nm as well as the minimum at 215 nm were found to be not
as intense in the presence of Ni2+ ions as without nickel
ions. The shoulder around 228 nm remains almost unaffected. These
changes suggest an increase in the amount of -turn in the secondary
structure (28).
PPIases were identified within all cell types and organisms, but little is known about their function in vivo. One commonly used strategy for investigation of protein function is the deletion or depletion of the appropriate gene followed by analysis of changes or defects of cellular functions. In E. coli, disruption of the gene encoding FKBP29 (fkbA) does not affect the phenotype (3). This may be due to functional redundancy of various proteins localized in the same cell compartment. This encouraged us to analyze two more proteins of E. coli with homology to FKBPs concerning their PPIase activity.
Chymotrypsin has been commonly used for isomer specific cleavage of the PPIase substrates in the protease coupled PPIase assay (1). Up to now, no proteolytic inactivation of PPIases was observed during the assay. This may be caused by insensitivity of most known PPIases to chymotrypsin. Another possibility is that the region responsible for PPIase activity remains intact if PPIases are proteolytically degraded during measurement (e.g. for E. coli trigger factor (29)).
For SlyD no PPIase activity was detected in the chymotrypsin-coupled
assay. Therefore, an assay with trypsin, another isomer specific
protease, was performed. On the basis of
kcat/Km the PPIase activity
of SlyD is considerably smaller than found for other PPIases with the
same substrate (for Suc-Ala-Ala-Pro-Arg-4NA: SlyD 6, 200 M1 s
1, trigger factor of
E. coli 499,000 M
1
s
1). The protein fragment representing amino acids 1-146
of SlyD was purified without proteolytic degradation and expressed a
quite similar kcat/Km
compared with the full-length SlyD. This confirms that amino acids
1-146 are an independent structural region of SlyD. The C-terminal
histidine-rich stretch of SlyD seems to be dispensable for PPIase
activity. However, the PPIase activity of SlyD was inhibited by
Ni2+ ions that probably bind to this region of the protein.
Because of the reversibility of this inhibition, we do not attribute it to chemical modification. Since the activity of the isolated N-terminal fragment of SlyD was not affected, binding of nickel ions to SlyD may
induce structural changes in the C-terminal domain, which in turn
influence the PPIase activity of the N-terminal domain. This mechanism
may be used in vivo to regulate the PPIase activity of SlyD.
SlyD is the only PPIase known so far that is specifically controlled by
high affinity metal ion binding. From CD spectra, alterations in the
secondary structure of SlyD by nickel ion binding were deduced. These
changes could be interpreted as an increase of the amount of
-turns.
A turn-stabilizing effect of cations bound to the backbone amide or
side chain functional groups has been occasionally observed for
peptides (30).
The C-terminal domain of SlyD shows homology to histidine-rich regions of GTPases involved in generation of active hydrogenases (HypB). Both the number of nickel ions bound per molecule SlyD (3) and the calculated Kd value for nickel cation binding to SlyD (1.8 µM) are in the same range as the values found in the literature for HypB. For HypB from Bradyrhizobium japonicum (24 clustered histidine residues), binding of 9 nickel ions with an Kd of 2.3 µM was detected (10); for HypB from Rhizobium leguminosarum (17 clustered histidine residues), 3.9 nickel ions/monomer with an Kd of 2.5 µM were detected (31). Hydrolysis of GTP by HypB has been found to be essential for nickel insertion in hydrogenases (32). Interestingly, HypB of E. coli lacks the histidine-rich region present at the N terminus of all other known HypB proteins and is not able to bind nickel but nevertheless exhibits GTPase activity (11). It is thus possible that SlyD plays the role of a nickel carrier in E. coli because its metal binding domain resembles those from HypB of other species. Thus SlyD could be involved in the generation of active hydrogenases in E. coli.
ORFs of H. influenzae, P. fluorescens, E. coli (ORF149), E. aerogenes, and M. jannaschii show high similarity to the N-terminal domain of SlyD
(83.6, 54.9, 53.4, and 47.3%, respectively). After detection of PPIase
activity for SlyD, similar activity was expected for the
SlyD-homologous protein products of the ORFs mentioned. To confirm
this, ORF149 of E. coli was cloned into an overexpression system. The purified protein SlpA showed a specificity constant kcat/Km of 7,400 M1 s
1 for the substrate
Suc-Ala-Phe-Pro-Arg-4NA. The substrate specificity of SlpA shows a
pattern different from other PPIases of E. coli (Table I).
The closest similarity was found to the E. coli trigger factor PPIase except for higher activity with substrates containing isoleucine or tryptophan at P1 position. By Western
blotting with antibodies directed against recombinant SlpA, the
protein product of ORF149 (SlpA) was now detected in E. coli. This was the first evidence for the expression of SlpA in
E. coli.
The alignment of amino acid sequences of SlyD, SlpA, trigger factor, and hFKBP12 (Fig. 1) shows that SlyD and SlpA are distinct from FKBPs and trigger factor though a certain degree of homology exists. SlyD and SlpA are 28.1 and 33.5% similar to hFKBP12 and 27.4 and 27.6% to trigger factor, respectively. The similarity between hFKBP12 and trigger factor was calculated to 37.1%. From the alignment and the comparison of specificity constants, the SlyD-homologous proteins seem to be as distinct as trigger factor from FKBPs.
Of the amino acids highly conserved in FKBPs, only a few are conserved
in SlyD and even less in SlpA (for review, see Ref. 33). Almost all
FK506 binding residues are substituted in SlpA, except for Ala-82,
Ile-92, and Phe-100. A Y27F mutation reduced the activity of hFKBP12 by
70% and binding of an FK506 homologue by 80% (34). This phenylalanine
residue is present in SlpA. Of special interest is the amino acid
substitution of Gly-29 in hFKBP12 to Val in SlyD and Leu in SlpA. These
rather bulky residues are inserted into the possible binding pocket for
FK506 (for hFKBP12, see Ref. 35). Moreover, both proteins as well as
all other SlyD-homologues contain a large insertion before the last
-strand compared with the structure of FKBP (33). This could
explain why the PPIase activity of SlyD and SlpA is not inhibited by
FK506. Because of this insensitivity and the distance to hFKBP12 and
trigger factor, but high sequence homology between the SlyD-homologous
proteins, we presume these proteins to be members of a new subfamily of FKBPs.
We are especially grateful to M. Schutkowski for synthesis of substrates and stimulating discussions. Further, we thank A. Schierhorn, for performing the mass spectrometry, and K. P. Rücknagel, for protein sequencing and HPLC.