PEDS Advance Access originally published online on July 14, 2006
Protein Engineering Design and Selection 2006 19(10):443-452; doi:10.1093/protein/gzl028
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Probing conserved amino acids in phospholipase D (Brassica oleracea var. capitata) for their importance in hydrolysis and transphosphatidylation activity
Martin-Luther University Halle-Wittenberg, Department of Biochemistry/Biotechnology Kurt-Mothes-Strasse 3, D-06120 Halle, Germany
1To whom correspondence should be addressed. E-mail: Ulbrich-Hofmann{at}biochemtech.uni-halle.de
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResults and discussionConclusionsAcknowledgementsReferences |
|---|
In addition to hydrolysis of glycerophospholipids, phospholipases D (PLDs) catalyze the head group exchange. The molecular basis of this transphosphatidylation potential, which strongly varies for PLDs from different sources, is unknown hitherto. Recently, the genes of two PLD isoenzymes from white cabbage have been sequenced and expressed in Escherchia coli, yielding the basis for mutational studies. In the present paper, three sequence characteristics of the isoenzyme (PLD2) that corresponds to the often used enzyme isolated from cabbage leaves have been probed for their importance in hydrolysis as well as transphosphatidylation activities: (i) the two HKD motifs, (ii) the C terminus and (iii) the eight cysteine residues. All these regions or amino acids are highly conserved in
-type plant PLDs. Based on multiple alignments, predictions of secondary structure and comparisons of hydrophobicity profiles, 35 enzyme variants were created and assayed. All positions tested proved to be very sensitive towards amino acid exchanges with respect to hydrolytic activity in the absence of glycerol as well as to the ratio of hydrolytic and transphosphatidylation activities in the presence of glycerol. A significant increase of total activity and transphosphatidylation activity could be obtained by the substitutions C310S and C625S.
Keywords: Brassica oleracea var. capitata/E.coli/phospholipase D/mutagenesis/transphosphatidylation
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionConclusionsAcknowledgementsReferences |
|---|
Phospholipase D (PLD; EC 3.1.4.4 [EC] ) occurring in plants, microorganisms and animals catalyzes the hydrolysis of glycerophospholipids at the terminal phosphodiester bond (Figure 1). In the presence of a suitable acceptor alcohol, most PLDs are able to transfer the phosphatidyl moiety to this alcohol in a so-called transphosphatidylation reaction (Figure 1) (Heller, 1978
). Because of this ability to catalyze the exchange of the head group alcohol in glycerophospholipids, PLD is used as biocatalyst in the transformation of phospholipids and phospholipid analogs in both laboratory and industrial scale (Servi, 1999; Ulbrich-Hofmann, 2000; Ulbrich-Hofmann et al., 2005). PLD from cabbage and PLDs from several Streptomyces species have been mostly used in biocatalytic transphosphatidylations, where enzymes from selected Streptomyces species (Streptomyces sp. strain PMF, Streptomyces antibioticus) seem to be superior over PLD from cabbage and other Streptomyces types (Juneja et al., 1988; Hirche and Ulbrich-Hofmann, 2000). An exact evaluation of different PLDs and general conclusions, however, are difficult, because the enzyme source, the acceptor alcohol and the solvent systems proved to be strongly interdependent (Hirche and Ulbrich-Hofmann, 2000). PLDs from plants are regarded as less appropriate for biosynthesis, although only a small number of their large variety have been tested so far. Thus, two new uncommon PLD isoenzymes from poppy have been described recently, one of which has a very high transphosphatidylation potential (Oblozinsky et al., 2003; Oblozinsky et al., 2005).
|
The molecular basis of the different transphosphatidylation propensities of PLDs is not known. Even though the crystal structure of one microbial PLD (from Streptomyces sp. strain PMF) has recently become accessible (Leiros et al., 2000
, 2004), structural features that are responsible for high or low ratios of transphosphatidylation to hydrolytic activity have not yet been found. The primary structures of plant PLDs show no significant similarity (<20%) with the amino acid sequence of this enzyme. Therefore, molecular modeling cannot be used to create an appropriate model of their tertiary structures. On the basis of the available primary structures, however, several common structural features could be ascertained. Despite the low identity between plant, microbial and human genes, several homologous regions could be derived (Figure 2). Sequence homologies have been found also with cardiolipin and phosphatidylserine synthases, several endonucleases, poxvirus envelope proteins and the murine toxin from Yersinia pestis, which resulted in the definition of the PLD superfamily (Ponting and Kerr, 1996). The most important common feature of the members of the superfamily is the presence of the so-called HKD motifs, corresponding to the conserved regions II and IV (Figure 2), comprising conserved His, Lys, Asp, Gly and/or Ser residues in a sequence of HxKxxxxDxxxxxxG(G/S). In PLD from Streptomyces sp. strain PMF they are located at the interface of the two structural domains of the monomeric PLD (Leiros et al., 2000). In most eukaryotic PLDs two additional conserved regions (regions I and III, Figure 2) with still unknown function were identified. Another typical structural feature of most plant PLDs is the N-terminal C2 domain (Figure 2), which is responsible for Ca2+-mediated phospholipid-binding (Nalefski and Falke, 1996; Wang, 2000).
|
Despite homology of all plant PLDs, according to their primary structures, several types of PLDs can be distinguished. In Arabidopsis thaliana, 12 isoenzymes have been found, which can be grouped into five PLD types (
-, ß-, 
-, 
-, and 
-type) (Qin and Wang, 2002). Most plant PLDs known so far belong to the -type.
In cabbage, the traditional and most often used source of plant PLD (Heller, 1978), we identified, sequenced and characterized two isoenzymes, PLD1 and PLD2, (Pannenberg et al., 1998; Schäffner et al., 2002). Both enzymes belong to the -type and have a sequence identity of 91%. The enzymes are monomers consisting of 810 and 812 amino acids. PLD isolated from cabbage leaves proved to be identical to PLD2 with respect to the amino acid sequence and the enzymatic properties but was modified by an N-terminal acetyl residue (Schöps et al., 2002). Therefore, the established recombinant cabbage PLD2 might be an appropriate system to localize amino acids responsible for different transphosphatidylation propensity in plant PLDs, which has not been studied hitherto.
In the present paper, several structural characteristics of PLD2, which are highly conserved in all plant PLDs, have been studied by site-directed mutagenesis to analyze their contribution to the hydrolytic and transphosphatidylation activity of the enzyme. Based on multiple alignments, predictions of secondary structure and comparisons of hydrophobicity profiles, 35 enzyme variants have been created and assayed. In probing the two HKD motifs, the C-terminus and the cysteine residues, most positions tested have proven to be sensitive towards amino acid exchanges with respect to hydrolytic and/or transphosphatidylation activity. For a few variants a significant increase in total activity and transphosphatidylation capacity could be observed.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionConclusionsAcknowledgementsReferences |
|---|
Materials
QIAquick Gel Extraction Kit and QIAprep Spin Miniprep Kit were from Qiagen (Hilden, Germany), QuikChangeTMSite-Directed Mutagenesis Kit from Stratagene Europe (Amsterdam, The Netherlands) and SequiTherm EXCEL II DNA Sequencing Kit was purchased from Biozym Diagnostik (Hess. Oldendorf, Germany). T4 DNA Ligase was from MBI Fermentas (St Leon-Rot, Germany) and the restriction enzymes were from New England Biolabs [Frankfurt (Main), Germany]. All mutagenic primers and IRD800-labeled primers were synthesized by MWG Biotech (Ebersberg, Germany). Escherichia coli XL1 blue was from Stratagene Europe (Amsterdam, The Netherlands), E.coli BL21(DE3), E.coli Rosetta(DE3)pLysS and pET-28b(+) were from Merck Biosciences (Bad Soden, Germany).
The expression construct pRSET5a-pld2 was generated by Schäffner et al. (2002). Dr U. Brinkmann (Xantos Biomedicine AG, München, Germany) supplied the vector pUBS520. Isopropyl-ß-D-thiogalactopyranosid (IPTG) was purchased from MBI Fermentas (St Leon-Rot, Germany), ampicillin and kanamycin from Serva (Heidelberg, Germany), chloroamphenicol from Sigma-Aldrich (München, Germany), tryptone as well as yeast extract from Becton Dickinson (Heidelberg, Germany) and non-fat milk from Oxoid (Basingstoke, UK).
BSA and bicinchoninic acid (BCA) were from Pierce (Rockford, USA). Anti-cabbage PLD antibodies were a gift from Dr J. Rajcani (Comenius University, Bratislava, Slovakia). Secondary peroxidase-conjugated antibodies and ECL-Plus Western Blotting Detection Kit were from Amersham Biosciences (Freiburg, Germany).
Phosphatidylcholine (PC), phosphatidic acid (PA) and phosphatidylglycerol (PG), all from soybean, were a gift from Lipoid GmbH (Ludwigshafen, Germany). Phosphatidyl-pnitrophenol (PpNP) was synthesized by Dr R. Schöps (Martin-Luther University Halle-Wittenberg, Germany) as described by D'Arrigo et al. (1995). Triton X-100 and high performance thin layer chromatography (HPTLC) plates (silica gel 60) were from Merck (Darmstadt, Germany). All other chemicals used were of analytical grade.
Bioinformatic analysis
Sequence analyses and alignments were carried out with Gene Runner (Hastings Software, Hasting, USA), Macaw (by Greg Schuler, NCBI Bethesda, USA) and BioEdit (by Tom Hall, Ibis Therapeutics Carlsbad, USA). Sequence data of PLDs from different sources were obtained from NCBI GenBank and UniProt. The hydrophobicity profiles were created by using BioEdit according to Kyte and Doolittle (1982). The secondary structures were predicted using the predictprotein server at the EMBL, Heidelberg, Germany, by PHD (http://cubic.bioc.columbia.edu/predictprotein; Rost, 1996).
Cloning of pld2 into pET-28b(+)
In the expression construct pRSET5a-pld2 the restriction site of NcoI within the pld2 gene was removed and a new NcoI restriction site necessary for cloning pld2 into pET-28b(+) was introduced by site-directed mutagenesis using the QuikChangeTMSite-Directed Mutagenesis Kit and the primer pairs 
ncopld2fw/ncopld2rv (5'-CTCGTGGCCAGGTTCATGGGTTCCGTAT-3') and mutncofw/mutncorv (5'-GAAGGAGATATACCCATGGCCCAGCA-3'). The resulting plasmid and the pET-28b(+) vector were digested with NcoI and HindIII, ligated and transformed into E.coli XL1 blue cells. After isolation of the plasmids, the correct nucleotide sequence of the complete insert was verified.
Construction of expression plasmids of PLD2 variants
The genes for the T264D-, Q334S-, K335R-, V339T/E342A/V343M-, D340E-, H663D-, S664H-, K665R-, D670E-, T812A-, T812C-, T812D-, T812F-, T812G-, T812K-, T812N-, T812R-, T812S-, T812V-, T812Y-, 808-812-, 804-812-, 801-812-, C85S-, C181S-, C212S-, C307S-, C310S-, C365S- and C741S-PLD2 variants were constructed by site-directed mutagenesis of the pld2 gene in the pRSET5a plasmid using the QuikChangeTMSite-Directed Mutagenesis Kit. The genes for the H333D-PLD2, V339T/E342A/V343M/Q346E/G348S/G349Q/S350Q/351/352/M356V-PLD2, 812-PLD2, 813S-PLD2 and C625S-PLD2 variants were constructed by site-directed mutagenesis of pld2 gene in pET-28b(+) plasmid using the QuikChangeTMSite-Directed Mutagenesis Kit. Plasmids were sequenced with the IRD800-labeled primers T7promotor (5'-CGAAATTAATACGACTCAC-3'), T7terminator (5'-GCTAGTTATTGCTCAGCGGTGG-3') and the gene specific primers PLD2midfw (5'-GTTCAGGACGTTGGACAC-3') and Sonde2rv (5'-TTACCACCYTGCTTTCTCCATCTCTGCTC-3') with a LiCor 4000 sequencer (MWG Biotech, Ebersberg, Germany). All constructs carrying the correct DNA sequences were transformed into E.coli BL21(DE3) containing the plasmid pUBS520 (Brinkmann et al., 1989) for the expression of pRSET5a-constructs or into E.coli Rosetta(DE3)pLysS for the expression of pET-28b(+)-constructs.
Expression of pld2 gene and its mutants in E.coli
The expression of the pld2 gene and its mutants in the E.coli expression host strain BL21(DE3) containing pUBS520 was performed as described for the non-modified recombinant PLD2 in Schäffner et al. (2002). The E.coli Rosetta(DE3)pLysS cells containing the appropriate pET-28b(+) expression construct were grown at 37°C in LB medium containing 10 µg ml–1 kanamycin and 25 µg ml–1 chloroamphenicol up to the late logarithmic phase. The cultures were diluted 1:50 with LB medium and grown at 37°C and 180 min–1 until OD600 reached 1. Expression of PLD2 and PLD2 variants was induced by the addition of IPTG to a final concentration of 25 µM. After induction, cultures were shaken for 4 h at 25°C and 180 min–1. Cells were harvested by centrifugation at 4°C and 5000 g for 10 min and stored at –80°C until needed.
Purification of PLD2 and its variants
The bacterial cells were resuspended in 30 mM piperazine-1,4-bis(2-ethanesulfonic acid) (PIPES) buffer, pH 6.2, containing 10 mM EDTA. Cells were disrupted by high-pressure homogenization (APV Homogeniser GmbH, Lübeck, Germany) and cellular debris was removed by centrifugation at 4°C and 30 000 g for 20 min.
For the purification of PLD2 and active enzyme variants expressed in E.coli BL21(DE3), CaCl2 was added to the crude extract for the subsequent Ca2+ mediated chromatography (final concentration 50 mM). The resulting solution was loaded onto an Octyl-Sepharose column (Amersham Biosciences, Freiburg, Germany) equilibrated with 30 mM PIPES, pH 6.2, 50 mM CaCl2. Proteins were eluted with 0.1 mM EDTA in 5 mM PIPES, pH 6.2. The eluates were analyzed for hydrolytic activity towards PpNP. After dialysis against 15 mM Tris/HCl, pH 7.5, the pooled active elution fractions were loaded onto a Q-Sepharose column (Amersham Biosciences, Freiburg, Germany) equilibrated with 15 mM Tris/HCl, pH 7.5. Proteins were eluted with 15 mM Tris/HCl, pH 7.5, 0.3 M NaCl. The active fractions were dialyzed against 5 mM PIPES, pH 6.2.
For the purification of PLD2 and active enzyme variants expressed in E.coli Rosetta(DE3)pLysS cells, the crude extract, adjusted to 15 mM PIPES, pH 7.5, was loaded onto a Q-Sepharose column equilibrated with 15 mM PIPES, pH 7.5. Proteins were eluted with 15 mM PIPES, pH 7.5, 0.2 M NaCl. CaCl2 was added to the collected and dialyzed active peak fractions (final concentration 50 mM) for the subsequent Ca2+ mediated chromatography. The pooled active fractions were loaded onto an Octyl-Sepharose column equilibrated with 30 mM PIPES, pH 6.2, 50 mM CaCl2. Proteins were eluted as described above.
For the partial purification of inactive PLD2 variants expressed in E.coli BL21(DE3) and E.coli Rosetta(DE3)pLysS, the crude extract was treated with (NH4)2SO4 (final concentration 0.4 M). After an incubation time of 1 h on ice, extracts were centrifuged at 21 000 g for 15 min. The resulting solution was dialyzed against 5 mM PIPES, pH 6.2. For purification of the inactive variants H333D- and D340E-PLD2 to homogeneity the crude extract was diluted 1:1 (v/v) with 30 mM PIPES pH 6.2, 0.8 M (NH4)2SO4 and loaded onto a Phenyl-Sepharose column equilibrated with 30 mM PIPES pH 6.2, 0.4 M (NH4)2SO4. Elution was performed with 30 mM PIPES pH 6.2. After dialysis against 20 mM Tris/HCl, pH 7.5, the pooled peak fractions were loaded onto a SOURCE Q column (Amersham Biosciences, Freiburg, Germany) equilibrated with 20 mM Tris/HCl, pH 7.5. Proteins were eluted with a linear gradient of 20 mM Tris/HCl, pH 7.5 and 20 mM Tris/HCl, pH 7.5, 2 M NaCl. The peak fractions were dialyzed against 10 mM PIPES, pH 7.0.
Protein determination
The protein content of the purified enzyme samples (PLD2, T264D-, V339T/E342A/V343M-, V339T/E342A/V343M/Q346E/G348S/G349Q/S350Q/351/352/M356V-, H333D-, D340E-, S664H-, T812N-, T812S-, T812C-, T812A-, T812V-, C85S-, C181S-, C212S-, C307S-, C310S-, C365S-, C625S-, and C741S-PLD2) was determined by the BCA assay with BSA as standard. The amount of protein in the partially purified samples (Q334S-, K335R-, H663D-, K665R-, D670E-, 812-, 808-812-, 804-812-801-812-, T812D-, T812K-, T812R-, T812F-, T812Y-, T812G- or 813S-PLD2) was determined by western blotting as follows. The protein samples were analyzed by SDS–PAGE according to Laemmli (1970) on 10% polyacrylamide gels and transferred to Hybond ECL nitrocellulose membranes (Amersham Biosciences, Freiburg, Germany). The blots were blocked with 4% non-fat milk and incubated with primary anti-cabbage PLD antibodies and with rabbit peroxidase-conjugated secondary antibodies. Immunoreactive bands were detected using the ECL-Plus Western Blotting Detection Kit. The scanned bands were quantified with AIDA software (raytest Isotopenmessgeräte GmbH, Straubenhardt, Germany). Defined amounts of purified PLD2 were used as standards.
CD and fluorescence spectra
Far-UV CD spectra of PLD2 and its variants (200 µg/ml) were recorded in 10 mM PIPES, pH 7.0, on a CD spectrometer J-810 (Jasco, Gross-Umstadt, Germany) at 20°C. The path length of the cuvette was 0.01 cm. Fluorescence spectroscopy was performed at a protein concentration of 10 µg/ml at 20°C with a FluoroMax-3 (Yvon-Spex, Grasbrunn, Germany) with excitation at 280 nm. The slit width was 5 nm for excitation and emission. 1 cm x 4 mm fluorescence cuvettes were used.
Enzyme activity assays in aqueous system with PpNP as substrate
The hydrolytic activities of PLD2 and PLD2 variants were determined during expression and purification processes in an aqueous micellar system by measurement of p-nitrophenol released from PpNP at 405nm in a microplate assay adapted to the procedure of D'Arrigo et al. (1995). The assay mixture consisted of 20 µl substrate solution (10 mM PpNP, 10% (v/v) Triton X-100 and 10 mM SDS), 170 µl 100 mM sodium acetate buffer, 50 mM CaCl2, pH 5.5 and 50 µl enzyme solution. After incubation for 20 min at 30°C, the reaction was stopped by addition of 60 µl 1 M Tris/HCl, pH 8.0, containing 0.1 M EDTA. The specific activity is expressed as amount of p-nitrophenol (µmol) released per minute and milligram protein.
No indication of a transphosphatidylation with Triton X-100 present in the assay was observed.
Enzyme activity assays in two-phase system with PC as substrate
The hydrolytic and transphosphatidylation activities of PLD2 and PLD2 variants were determined in a two-phase system according to Hirche et al. (1997). 620 µl of n-hexane containing the substrate PC (1.3 µmol) and 2-octanol (0.23 mmol) were placed into 1.5 ml vials. The reaction was started by the injection of 80 µl buffer (300 mM sodium acetate, 120 mM CaCl2, pH 5.6) containing PLD2 or PLD2 variants. In the case of transphosphatidylation reactions, the buffer additionally contained the acceptor alcohol glycerol (108 µmol). The conversions were performed on a horizontal shaker at 30°C and 300 min–1. Aliquots of the organic phase were withdrawn at several periods of time and analyzed for phospholipid content by HPTLC with densitometric evaluation of the spots after staining with cupric sulfate/phosphoric acid (Touchstone et al., 1983) using standards of PC, PA and PG. The aqueous phase did not contain detectable amounts of phospholipids. The initial rates of hydrolysis or transphosphatidylation were obtained from the linear fit of the amount of reaction products as a function of reaction time. All results are the means of duplicates of two independent determinations. No indication of a transphosphatidylation with 2-octanol present in the assay was observed.
|
Results and discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionConclusionsAcknowledgementsReferences |
|---|
Design of PLD2 variants
Tailoring the activity and specificity of enzymes is one of the most challenging objectives of protein engineering. With PLD, increasing the ratio of hydrolysis and transphosphatidylation activities (Figure 1) in favor of transphosphatidylation is desired for the biotransformation of phospholipids. Moreover, exploring the structural reasons for the different transphosphatidylation potentials of PLDs originating from different sources is of basic interest. Our comparison of the primary sequences of PLDs from cabbage with those of other PLDs revealed some distinctive features that were probed by the construction and investigation of 35 variants of PLD2 from cabbage. For all variants with more than one amino acid exchange, a prediction of secondary structure was performed without any indication that the content of helix (20%), sheet (25%) and other structural elements (55%) were significantly changed in comparison with the secondary structure of the wildtype PLD2 (Figure 3).
|
Modifications in and near the HKD motifs. Although the two HKD motifs are found in all PLDs, the first (N-terminal) HKD motif (region II in Figure 2) seems less conserved than the second (C-terminal) one (region IV in Figure 2), as a multiple sequence alignment showed (Figure 4A). Because the two HKD motifs probably form the active site as found for microbial PLD (Leiros et al., 2000
), differences in the transphosphatidylation potential are expected to be related to the sequence of the first HKD motif. Starting from this assumption, several variants derived from the sequences of PLDs with striking differences in these regions were created. The single variant T264D-PLD2, the threefold variant V339T/E342A/V343M-PLD2 and the tenfold variant V339T/E342A/V343M/Q346E/G348S/G349Q/S350Q/351/352/M356V-PLD2 were derived from the sequence of poppy PLD1 (Figure 4A). Interestingly, two PLD isoenzymes isolated from poppy seedlings had shown extremely different hydrolysis and transphosphatidylation activities (Oblozinsky et al., 2003), whereas recombinantly produced PLD1 and PLD2 from poppy seem to be different from these enzymes (Lerchner et al., 2005).
|
The variants Q334S-PLD2 and S664H-PLD2 with mutations between the presumably essential His and Lys residues in the first or second HKD motif were derived from the sequence comparison of PLD2 from cabbage and PLD from Streptomyces species (Figure 4B). The HKD motifs are the only regions of the primary structure of plant and Streptomyces PLDs that show significant similarity. This comparison is of particular interest, because the tertiary structure of PLD from Streptomyces sp. strain PMF is known (Leiros et al., 2000
, 2004), and this enzyme has a high transphosphatidylation activity (Carrea et al., 1995).
Finally, the variants H333D-, K335R-, D340E-PLD2 and the variants H663D-, K665R-, D670E-PLD2 were designed to check the necessity of these highly conserved amino acid residues in the HKD motifs for the catalytic activity of the plant -type PLDs. These mutations were selected in analogy to the structural analysis of human PLD1 (Sung et al., 1997). While the effect of charge is checked by the substitution of His by Asp, the influence of spacial orientation and distances is probed by the substitutions of Lys by Arg and Asp by Glu residues.
Modifications at the conserved C-terminus.
Mammalian PLDs possess a highly conserved C-terminus with the sequence PMEVWT required for catalytic activity as concluded from mutational studies on rat brain PLD1 (Xie et al., 2000) and human PLD1 (Liu et al., 2001). Our alignment studies indicated the presence of a highly conserved C-terminus also for plant PLDs of the -type with the sequence PPILTT (Figure 4C). In contrast, the C-terminus of microbial PLDs is not conserved. Therefore, we created several deletion variants (801-812-PLD2, 804-812-PLD2, 808-812-PLD2, 812-PLD2) to probe the importance of the C-terminal amino acids in PLD2 for hydrolysis and transphosphatidylation activity. Moreover, a variant with an additional C-terminal Ser residue (813S-PLD2) and several variants with a substitution in the amino acid position 812 (T812D-, T812K-, T812R-, T812F-, T812Y-, T812G-, T812N-, T812S-, T812C-, T812A-, T812V-PLD2) were designed to examine the influence of the type of C-terminal amino acid on the activities.
Substitution of the conserved eight cysteine residues.
Plant PLDs of the -type as well as Streptomyces PLDs possess eight highly conserved cysteine residues. While the cysteine residues of the intracellular plant PLDs are present in reduced form as shown for PLD from cabbage (Hwang et al., 2001), these residues seem to form disulfide bonds in extracellular PLDs as can be concluded from the crystal structure of PLD from Streptomyces sp. strain PMF (PDB accession No. 1V0S) and studies on folding of PLD from S.antibioticus (Iwasaki et al., 2000). To probe a potential function of the eight highly conserved cysteine residues in plant PLDs of the -type, the individual cysteine residues of PLD2 (Figure 5) were exchanged by serine residues, yielding the enzyme variants C85S-PLD2, C181S-PLD2, C212S-PLD2, C307S-PLD2, C310S-PLD2, C365S-PLD2, C625S-PLD2 and C741S-PLD2. Serine residues were chosen for this substitution, because they are most similar to cysteine residues and will preserve hydrogen bonds.
|
Expression and purification of PLD2 variants
The expression of pld2 and pld2 variants was performed using the pRSET5a-vector and the E.coli strain BL21 (DE3)+pUBS520 or the pET-28b(+)-vector and the E.coli strain Rosetta(DE3)pLysS, respectively, as described in Materials and methods. The introduction of the second expression strategy became necessary because of an observed plasmid instability of PLD2 in the BL21(DE3) system, particularly to ensure the results of the inactive variants. After disintegration of the harvested cells, the crude extract of the active variants was purified by Ca2+ ion mediated hydrophobic interaction chromatography on Octyl-Sepharose and anion-exchange chromatography on Q-Sepharose. In case of the variants expressed in the E.coli strain Rosetta(DE3)pLysS a preceding anion-exchange chromatography on Q-Sepharose was necessary to remove interfering components of the bacterial culture. All the active PLD2 variants obtained by any of these two protocols were homogeneous in SDS–PAGE.
Interestingly, variants showing no activity in the crude extract could not be purified by the Ca2+ ion mediated hydrophobic interaction chromatography on Octyl-Sepharose. This procedure, which has proven very efficient in purifying plant PLDs (Lambrecht and Ulbrich-Hofmann, 1992; Abousalham et al., 1993; Novotna et al., 1999; Oblozinsky et al., 2003), is based on the binding of PLD onto Octyl-Sepharose in the presence of Ca2+ ions and the subsequent selective elution of the enzyme by EDTA. As shown recently for PLDs from poppy, the affinity of PLD to the carrier correlates with its activity (Lerchner et al., 2005) explaining the failure of the method in the purification of inactive PLD2 variants. To check the hydrolytic and transphosphatidylation activities of the inactive variants in the two-phase system as described below, these enzymes were purified only partially by (NH4)2SO4 precipitation. The protein content of these enzymes was determined by western blotting using standards of PLD2 for calibrations as described in Materials and methods. A purification to homogeneity in SDS–PAGE was achieved for H333D- and D340E-PLD2 by two steps using hydrophobic interaction (Phenyl Sepharose) and anion-exchange chromatography (Source-Q) (Figure 6). The resulting proteins showed the same far-UV CD-spectra as non-modified PLD2 (Figure 7A). Interestingly, fluorescence spectra (Figure 7B) show a shifting of the fluorescence maximum for the inactive variants. Obviously, the secondary structure of PLD2 is completely conserved in H333D- and D340E-PLD2, whereas there are certain changes in their tertiary structures. Further studies on the folding of PLD2 and selected variants are in progress.
|
|
Hydrolytic and transphosphatidylation activities of PLD2 variants with modifications in the HKD motifs
The hydrolytic and transphosphatidylation activities of PLD2 and its variants with single, threefold or tenfold mutations in or in close vicinity of the two HKD motifs were compared in the conversion of PC from soya in a two-phase system consisting of n-hexane/2-octanol and sodium acetate buffer. The reaction was performed in the absence of glycerol yielding the hydrolytic activity (vH) as well as in the presence of glycerol vT, where the ratio of the initial rates of transphosphatidylation and hydrolysis (vT/vH) serves as a measure of the transphosphatidylation propensity (Table I).
|
The complete loss of activity of the variants H333D-, K335R-, D340E-, H663D-, K665R- and D670E-PLD2 (Table I) proves that the amino acid residues His, Lys and Asp in the two HKD motifs (Figure 2), which are typical for all members of the PLD superfamily, are absolutely essential for the catalytic activity of
-type plant PLDs, too. Assuming that the mechanism of catalysis by PLD2 (cabbage) is similar to that suggested for PLD from Streptomyces sp. PMF (Leiros et al., 2004), the nucleophilic attack of the imidazole-N of His333 on the phosphorus atom of the substrate is prevented in H333D-PLD2 variant, while the missing His residue in H663D-PLD2 impedes the protonation of the substrate oxygen and, therefore, the release of the polar head group. The Lys residues 335 and 665, corresponding to Lys172 and Lys450 in PLD from Streptomyces sp. PMF (Leiros et al., 2004), probably interact with the oxygen atoms of the phosphate residue in the active site. Interestingly, the hydrolytic and transphosphatidylation activities of the Lys and Asp mutants were completely abolished, although the positive charge in the enzyme variants K335R- and K665R-PLD2 or the negative charge in D340E- and D670E-PLD2 was preserved. Obviously, small changes in the orientation and distances of the charges in these positions are sufficient to disturb the catalytic mechanism. The essential Asp residues 340 and 670 are probably involved in the catalysis in an indirect way by stabilizing the active site conformation or by contributing to the interfacial binding of the enzyme to the substrate as can be concluded from the positions of the corresponding Asp residues 177 and 455 in PLD from Streptomyces sp. PMF, which are relatively remote from the active site. The results of the enzyme variants Q334S- and S664H-PLD2, which were modified between the essential His and Lys residues of the first or second HKD motif (Figure 4B), indicate a great influence of the substitutions in these positions on activity (Table I). While the mutation Q334S resulted in total loss of activity, the enzyme variant S664H showed an increase in the hydrolytic activity in the absence of glycerol and a higher transphosphatidylation activity compared with the wildtype enzyme (Table I). Therefore, these positions, particularly position 664, might be appropriate candidates for further mutational studies with respect to increasing the transphosphatidylation potential of plant as well as of microbial PLDs.
The hydrolytic and transphosphatidylation activities of the enzyme variants T264D-PLD2, V339T/E342A/V343M-PLD2 and V339T/E342A/V343M/Q346E/G348S/G349Q/S350Q/351/352/M356V-PLD2, which were derived from the sequence of the poppy PLDs, revealed an increase in the transphosphatidylation potential (vT/vH) in all cases (Table I). The effects were, however, not dramatic and support the finding that neither of the recombinant PLDs from poppy (Lerchner et al., 2005) is identical with one of the two uncommon PLDs isolated from poppy seedlings (Oblozinsky et al., 2003). One of these enzymes possesses no transphosphatidylation activity and has an optimum hydrolytic activity at pH 8.0, whereas the other one has an extremely high transphosphatidylation preference and an activity optimum at pH 5.5.
Hydrolytic and transphosphatidylation activities of PLD2 variants with modifications at the C-terminus
The results of the mutations at the C-terminus indicate that the highly conserved C-terminal amino acid residues of the -type plant PLDs are essential for enzymatic activity (Table II). The deletion of the last amino acid residue (Thr812) in the enzyme variant 812-PLD2 is sufficient for the complete loss of activity. The activity was also completely abolished if the C-terminal Thr was replaced by charged amino acid residues in the variants T812D-, T812K- and T812R-PLD2 by aromatic amino acid residues in the variants T812F- and T812Y-PLD2 or by the small Gly residue in the variant T812G-PLD2. Also the insertion of an additional amino acid residue (813S-PLD2) resulted in complete loss of activity. These results suggest that the C-terminus of PLD2 is hidden inside the protein molecule.
|
The substitution of Thr812 by polar, uncharged amino acid residues as in T812N-, T812S- and T812C-PLD2 preserves some activity. Interestingly, the hydrolytic activity is decreased more than the transphosphatidylation activity, whereby the ratio vT/vH is increased in comparison with wildtype PLD2 (Table II). In contrast, the insertion of the apolar, aliphatic amino acid residues Ala and Val in T812A- and T812V-PLD2 did not significantly change the ratio vT/vH but showed different effects on total activity. Marked differences were observed in the absence of glycerol, where T812A-PLD2 caused a dramatical decrease in hydrolytic activity, but the activity of T812V-PLD2 was more than doubled in comparison with wildtype PLD2 (Table II).
The presented results suggest that the C-terminus contributes to the functional conformation of active -type plant PLDs, probably through hydrophobic interactions, similarly as Liu et al. (2001) proposed for mammalian PLDs. Furthermore, it seems that modification of the C-terminus allows to influence activity in favor of transphosphatidylation.
Hydrolytic and transphosphatidylation activities of PLD2 variants with substitutions of the cysteine residues
In no case did the exchange of any cysteine residue in PLD2 by a serine residue result in a complete loss of hydrolytic or transphosphatidylation activity (Table III), suggesting that the tertiary structure of the enzyme is preserved. The activity of enzyme variants C85S- and C307S-PLD2, however, was strongly decreased. As concluded from the secondary structure prediction (Figure 3), the Cys residues 85 and 307 are integrated into ß-sheet structures. The variants C365S- and C741S-PLD also showed decreased activities compared with the wildtype enzyme. In the glycerol-containing reaction system, the hydrolytic reaction was more strongly decelerated than the transphosphatidylation reaction resulting in a higher ratio vT/vH (Table III). This trend was still more distinct in the enzyme variants C181S- and C212S-PLD2 where the decrease in transphosphatidylation activity was small or not present in contrast to the hydrolytic activity. A significant increase in the total activity with a distinct preference for transphosphatidylation activity, finally, was observed for the variants C310S- and C625S-PLD2 (Table III). Interestingly, the mutations in these variants are, based on the primary structure, near to the two catalytic HKD motifs (Figure 5) and surrounded by highly conserved amino acid residues (Figure 5). In the enzyme variant C625S-PLD2 the ratio vT/vH could be increased by the factor of 2.8 (Table III).
|
The results suggest that all the highly conserved cysteine residues are involved in the preservation of local structural elements of the enzyme and/or in the binding of the acceptor molecules (alcohol and water).
Interestingly, the cysteine residues of PLD from Streptomyces species strain PMF, which form four disulfide bridges in contrast to the free cysteine residues in PLD2, are located in loop-regions at the surface, as can be concluded from the crystal structures of PLD from Streptomyces species strain PMF (PDB Accession-No 1V0S and 1V0W; Leiros et al., 2004).
In PLD2 from cabbage the two cysteine residues near the two HKD motifs might be directly involved in the binding of the acceptor molecules and should be, therefore, appropriate starting points for the improvement of the transphosphatidylation potential by site-directed mutagenesis. Further kinetic studies, however, will be necessary to draw unambiguous conclusions.
In all our experiments dithiothreitol had no significant effect on the activity of the variants and no oligomer formation was observed in native PAGE.
|
Conclusions |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionConclusionsAcknowledgementsReferences |
|---|
The present studies, which are—to the best of our knowledge—the first mutational studies on a plant PLD and the first studies on the transphosphatidylation potential of PLD, show that the two His, Lys and Asp residues in the HKD motifs are absolutely essential for hydrolytic as well as transphosphatidylation activity. Moreover, the occupancy of the C-terminal position 812 proved to be essential. The eight conserved cysteine residues influence activities to different extents. Modification of the position 664 in the second HKD motif was shown to allow an increase in the hydrolytic activity. Moreover, the amino acid residue in the C-terminal position 812 and the two cysteine residues near the HKD motifs (C310 und C625) yield possible scope to increase the transphosphatidylation potential of PLD2. Further insights, however, will need a kinetic analysis of the hydrolysis and transphosphatidylation reactions, which is not trivial in two-phase systems because of partition effects and the supramolecular substrate structure.
|
Footnotes |
|---|
Edited by Andreas Kungl
|
Acknowledgements |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionConclusionsAcknowledgementsReferences |
|---|
We thank Dr I. Schäffner (First Department of Internal Medicine, Martin-Luther University Halle-Wittenberg) for the generation of the expression construct pRSET5a-pld2, Dr R. Schöps and Ch. Kuplens (Institute of Biotechnology, Martin-Luther University Halle-Wittenberg) for the preparation of phosphatidyl-p-nitrophenol and help in the purification of the PLD2 variants. We would also like to thank Dr J. Rajcani (Comenius University Bratislava, Slovakia) for providing the anti-cabbage PLD antibodies, Dr U. Brinkmann (Xantos Biomedicine AG, München, Germany) for kindly supplying the vector pUBS520 and the Lipoid GmbH (Ludwigshafen) for the gift of the phospholipids. The support for this work by a grant from the Federal State of Saxony-Anhalt for A. Lerchner is gratefully acknowledged.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionConclusionsAcknowledgementsReferences |
|---|
Abousalham A., Riviere M., Teissere M., Verger R. (1993) Biochim. Biophys. Acta 1158:1–7.[Medline]
Brinkmann U., Mattes R.E., Buckel P. (1989) Gene 85:109–114.[CrossRef][Web of Science][Medline]
Carrea G.D., D'Arrigo P., Piergianni V., Roncaglio S., Secundo F., Servi S. (1995) Biochim. Biophys. Acta 1255:273–279.[Medline]
D'Arrigo P., Piergianni V., Scarcelli D., Servi S. (1995) Anal. Chim. Acta 304:249–254.
Heller M. (1978) Adv. Lipid Res. 16:267–326.[Medline]
Hirche F., Koch M.H.J., König S., Wadewitz T., Ulbrich-Hofmann R. (1997) Enzyme Microb Technol. 20:453–461.
Hirche F. and Ulbrich-Hofmann R. (2000) Biocatal. Biotransformation 18:343–353.
Hwang I.-S., Park S.-J., Roh T.-Y., Choi M.-U., Kim H.J. (2001) Rapid Commun Mass Spectrom. 15:110–115.[CrossRef][Web of Science][Medline]
Iwasaki Y., Nishiyama T., Kawaraski Y., Nakano H., Yamane T. (2000) J. Biosci. Bioeng. 89:506–508.[CrossRef][Web of Science][Medline]
Juneja L.R., Kazuoka T., Yamane T., Shimizu S. (1988) Biochim. Biophys. Acta 960:334–341.[Medline]
Kyte J. and Doolittle R.F. (1982) J. Mol. Biol. 157:105–132.[CrossRef][Web of Science][Medline]
Laemmli U.K. (1970) Nature 227:680–685.[CrossRef][Medline]
Lambrecht R. and Ulbrich-Hofmann R. (1992) Biol. Chem. Hoppe-Seyler 373:81–88.[Web of Science][Medline]
Leiros I., Secundo F., Zambonelli C., Servi S., Hough E. (2000) Structure 8:655–667.[Medline]
Leiros I., McSweeney S., Hough E. (2004) J. Mol. Biol. 339:805–820.[CrossRef][Web of Science][Medline]
Lerchner A., Mansfeld J., Schäffner I., Schöps R., Beer H.K., Ulbrich-Hofmann R. (2005) Biochim. Biophys. Acta 1737:94–101.[Medline]
Liu M.Y., Gutowski S., Sternweis P.C. (2001) J. Biol. Chem. 276:5556–5562.
Nalefski E.A. and Falke J.J. (1996) Protein Sci 5:2375–2390.[Web of Science][Medline]
Novotna Z., Kas J., Daussant J., Sajdok J., Valentova O. (1999) Plant Physiol. Biochem. 37:531–537.
Oblozinsky M., Schöps R, Ulbrich-Hofmann R., Bezakova L. (2003) Biochim. Biophys. Acta 1631:153–159.[Medline]
Oblozinsky M., Ulbrich-Hofmann R., Bezakova L. (2005) Biotechnol. Lett. 27:181–185.[CrossRef][Web of Science][Medline]
Pannenberg I., Mansfeld J., Ulbrich-Hofmann R. (1998) Plant Physiol. 118:1102.
Ponting C.P. and Kerr I.D. (1996) Protein Sci. 5:914–922.[Web of Science][Medline]
Qin C. and Wang X. (2002) Plant Physiol. 128:1057–1068.
Rost B. (1996) Methods Enzymol. 266:525–539.[CrossRef][Web of Science][Medline]
Schäffner I., Rücknagel K.-P., Mansfeld J., Ulbrich-Hofmann R. (2002) Eur. J. Lipid Sci. Technol. 104:79–87.[CrossRef]
Schöps R., Schierhorn A., Schäffner I., Mansfeld J., Ulbrich-Hofmann R. (2002) J. Protein Chem. 21:407–411.[CrossRef][Web of Science][Medline]
Servi S. (1999) Topics Curr. Chem. 200:127–158.
Sung T.C., Roper R.L., Zhang Y., Temel R., Hammond S.M., Morris A.J., Moss B., Engebrecht J., Frohman M.A. (1997) EMBO J 16:4519–4530.[CrossRef][Web of Science][Medline]
Touchstone J.C., Levin S.S., Dobbins M.F., Beers P.C. (1983) J. Liq. Chromatogr. 6:179–192.
Ulbrich-Hofmann R. (2000) In Bornscheuer U.T. (Ed.). Enzymes in Lipid Modification (Wiley-VCH, Weinheim) pp. 219–262.
Ulbrich-Hofmann R., Lerchner A., Oblozinsky M., Bezakova L. (2005) Biotechnol. Lett 27:535–543.[CrossRef][Web of Science][Medline]
Wang X. (2000) Prog. Lipid Res 39:109–149.[CrossRef][Web of Science][Medline]
Xie Z., Ho W.T., Exton J.H. (2000) Eur. J. Biochem 267:7138–7146.[Web of Science][Medline]
Received February 14, 2006; revised May 19, 2006; accepted June 12, 2006.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||