Glycobiology Advance Access originally published online on April 20, 2005
Glycobiology 2005 15(9):874-886; doi:10.1093/glycob/cwi066
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Processive lipid galactosyl/glucosyltransferases from Agrobacterium tumefaciens and Mesorhizobium loti display multiple specificities
2 Biozentrum Klein Flottbek, University of Hamburg, Hamburg 22609, Germany; and 3 Department of Immunochemistry and Biochemical Microbiology, Research Center Borstel, Leibniz-Center for Medicine and Biosciences, Borstel 23845, Germany
1 To whom correspondence should be addressed; e-mail: warnecke{at}botanik.uni-hamburg.de
Received on March 11, 2005; revised on April 12, 2005; accepted on April 16, 2005
 |
Abstract |
|---|
 Top Abstract IntroductionResultsDiscussionMaterials and methodsSupplementary dataAcknowledgmentsReferences |
|---|
The glycosyltransferase family 21 (GT21) includes both enzymes of eukaryotic and prokaryotic organisms. Many of the eukaryotic enzymes from animal, plant, and fungal origin have been characterized as uridine diphosphoglucose (UDP-Glc):ceramide glucosyltransferases (glucosylceramide synthases [Gcs], EC 2.4.1.80 [EC] ). As the acceptor molecule ceramide is not present in most bacteria, the enzymatic specificities and functions of the corresponding bacterial glycosyltransferases remain elusive. In this study, we investigated the homologous and heterologous expression of GT21 enzymes from Agrobacterium tumefaciens and Mesorhizobium loti in A. tumefaciens, Escherichia coli, and the yeast Pichia pastoris. Glycolipid analyses of the transgenic organisms revealed that the bacterial glycosyltransferases are involved in the synthesis of mono-, di- and even tri-glycosylated glycolipids. As products resulting from their activity, we identified 1,2-diacyl-3-(O-b-d-galacto-pyranosyl)-sn-glycerol, 1,2-diacyl-3-(O-b-d-gluco-pyranosyl)-sn-glycerol as well as higher glycosylated lipids such as 1,2-diacyl-3-[O-b-d-galacto-pyranosyl-(1
6)-O-b-d-galacto-pyranosyl]-sn-glycerol, 1,2-diacyl-3-[O-b-d-gluco-pyranosyl-(16)-O-b-d-galacto-pyranosyl]-sn-glycerol, 1,2-diacyl-3-[O-b-d-gluco-pyranosyl-(16)-O-b-d-gluco-pyranosyl]-sn-glycerol, and the deviatingly linked diglycosyldiacylglycerol 1,2-diacyl-3-[O-b-d-gluco-pyranosyl-(13)-O-b-d-galacto-pyranosyl]-sn-glycerol. From a mixture of triglycosyldiacylglycerols, 1,2-diacyl-3-[O-b-d-galacto-pyranosyl-(16)-O-b-d-galacto-pyranosyl-(16)-O-b-d-galacto-pyranosyl]-sn-glycerol could be separated in a pure form. In vitro enzyme assays showed that the glycosyltransferase from A. tumefaciens favours uridine diphosphogalactose (UDP-Gal) over UDP-Glc. In conclusion, the bacterial GT21 enzymes differ from the eukaryotic ceramide glucosyltransferases by the successive transfer of up to three galactosyl and glucosyl moieties to diacylglycerol. Key words: galactosyl diacylglycerol / galactosyltransferase / GCS / glucosylceramide / glycosyltransferase family 21
|
Introduction |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsSupplementary dataAcknowledgmentsReferences |
|---|
Glycoconjugates, in the plant and animal kingdom as well as in prokaryotic cells, are formed by the action of a multiplicity of different glycosyltransferases. At present, several thousands of these glycosyltransferases are classified into 77 distinct sequence-based families (Campbell et al., 1997
; Coutinho et al., 2003; Carbohydrate-Active Enzymes server at URL: http://afmb.cnrs-mrs.fr/CAZY/). The value of this classification continuously increases by the biochemical characterization of additional members of each of these glycosyltransferase families. Such characterization has been performed with many eukaryotic representatives of the glycosyltransferase family 21 (GT21) (Ichikawa et al., 1996; Ichikawa and Hirabayashi, 1998; Wu et al., 1999; Leipelt et al., 2001; Hillig et al., 2003; Kohyama-Koganeya et al., 2004). All GT21 genes studied so far from animal, fungal, and plant origin have been found to code for uridine diphosphoglucose (UDP-Glc):ceramide glucosyltransferase (synonymous with glucosylceramide synthase [Gcs], EC 2.4.1.80
[EC]
). The formation of glucosylceramide (GlcCer) represents the first glycosylation step in the biosynthesis of more complex glycosphingolipids which occur ubiquitously in cell membranes of eukaryotic organisms (Warnecke and Heinz, 2003).
Glycosphingolipids are also found in a few bacterial genera such as Sphingobacterium and Sphingomonas (Kawahara et al., 1991, 2000; Olsen and Jantzen, 2001). Therefore, it does not seem surprising that GT21 includes also members of bacterial origin. Most of these hypothetical proteins had been referred to as Gcs by automatic open reading frame (ORF) annotation, although none of them has been characterized experimentally up to now. Bacterial representatives containing a putative gcs+ gene are Agrobacterium tumefaciens and Mesorhizobium loti. The plant parasite A. tumefaciens belongs to the family of Rhizobiaceae with members able for fixing nitrogen and maintaining symbiosis with plants. The nitrogen-fixing plant symbiont M. loti belongs to the closely related family of Phyllobacteriaceae (Young et al., 2001). Interestingly, the occurrence of glycosylceramides or other glycosphingolipids has never been reported for members of Rhizobiaceae and for M. loti (Wilkinson, 1988). With the exception of Rhizobium (Orgambide et al., 1992), not even glycoglycerolipids, which predominantly occur in Gram-positive bacteria and cyanobacteria, have been found in these organisms. Because of the lack of such glycolipids, it is surprising that both A. tumefaciens and M. loti contain putative gcs+ genes. In view of these facts, the actual enzymatic activity of bacterial representatives of GT21 remains obscure.
The aim of this study was to determine the enzymatic function of the hypothetical Gcs enzymes from A. tumefaciens and M. loti. For this purpose, we analyzed the lipid extract of A. tumefaciens for the occurrence of glycosylceramides and glycoglycerolipids. But because these glycolipids could not be detected, both bacterial Gcs sequences were cloned and expressed in different host organisms with subsequent glycolipid analyses and enzyme assays. In the following, we will demonstrate that the Gcs from A. tumefaciens and M. loti differ from the eukaryotic members of GT21 in several ways: they favour uridine diphosphogalactose (UDP-Gal) over UDP-Glc as sugar donor and diacylglycerol (DAG) over ceramide as acceptor. In addition, the bacterial Gcs consecutively transferred up to three glycosyl residues to an accordingly changed lipid acceptor.
|
Results |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsSupplementary dataAcknowledgmentsReferences |
|---|
Cloning of glycosyltransferases from Agrobacterium tumefaciens and Mesorhizobium loti
Our efforts to find the ORFs encoding putative glycosyltransferases in the two bacteria made use of the BLAST search (Altschul et al., 1990)
based on the amino acid sequence of the Gcs from Homo sapiens (Ichikawa et al., 1996). Two promising candidates were recognized in the fully sequenced genomes of A. tumefaciens (GenBank accession number, NP_354792
[GenBank]
; locus tag, AGR_C_3323) and M. loti (GenBank accession number: NP_106273
[GenBank]
; locus tag, mlr5650). The two corresponding polypeptides shared 60% identity, whereas their similarity to the query sequence was significantly lower (about 23%). The recently characterized Gcs enzymes from fungi and plants showed similarities even as low as 9–21 % to the human Gcs (Leipelt et al., 2001).
Due to sequence similarity, the encoded enzymes from A. tumefaciens and M. loti fall into the glycosyltransferase family GT21 (Campbell et al., 1997; Coutinho et al., 2003; Carbohydrate-Active Enzymes server at URL: http://afmb.cnrs-mrs.fr/CAZY/) (Figure 1A). Like all other members of the family, they contain a putative N-terminal transmembrane domain and the widely spaced D1,D2,D3(Q/R)XXRW motif (Marks et al., 2001) (Figure 1B). This motif was previously shown to be characteristic for processive ß-glycosyltransferases of GT2 (Saxena et al., 1995). Although D1,D2,D3(Q/R)XXRW is present in all Gcs sequences (Leipelt et al., 2001; Marks et al., 2001), none of the GT21 family members which have been characterized so far, showed processivity.
|
On the basis of the sequence similarity to the other Gcs enzymes of GT21, the sequences from A. tumefaciens and M. loti had been automatically annotated as putative ceramide glucosyltransferases. However, enzymatic evidence for their functions has not been provided so far, and neither the actual sugar donors nor the sugar acceptors were known. Therefore, we tried to identify the enzymatic activity of the two enzymes by various approaches based on the assumption that the glycosyltransferases may contribute to the biosynthesis of glycolipids.
Agrobacterium tumefaciens lacks detectable proportions of "conventional" glycolipids
As a first approach to identify the function of the glycosyltransferase from A. tumefaciens, we prepared lipid extracts from this bacterium and looked for the presence of conventional glycolipids extractable by chloroform /methanol. Thin layer chromatography (TLC) of the total lipid extract did not show the presence of any glycolipids. To exclude that very low proportions of glycolipids escaped detection, we fractionated the total lipid extract by preparative column chromatography into neutral lipids, glycolipids, and phospholipids. All three fractions were redissolved in very small volumes of solvent for subsequent spotting onto TLC plates, but still we could not detect any conventional glycolipids in any of the three fractions (Figure 2A). To find out, whether the glycosyltransferase gene of A. tumefaciens exerts any effect on the lipid pattern of this bacterium, we deleted this gene by homologous recombination with an antibiotic resistance cassette (Figure 2C). The lipid extract of the transformed bacteria was subjected to the procedure described above, but no change compared to wild type cells, and again no glycolipids could be detected (Figure 2A). From these results, we conclude that the absence of the putative glycosyltransferase activity and its products are not lethal for A. tumefaciens.
|
Finally, we checked the possibility that a product of the Gcs glycosyltransferase activity was not detectable, because of promoter repression under the growth conditions used. For this purpose, the ORF of the gcs+ was replaced by a heterologous glucosyltransferase from Staphylococcus aureus (Ugt106B1), serving as a reporter sequence to determine the activity of the gcs+ promoter (Figure 2D). We had shown before (Jorasch et al., 2000) that the enzyme encoded by ugt106B1 on expression in various hosts resulted in the formation of the diglucosyldiacylglycerol ßGlcßGlcD (abbreviations and structures are given in Table I). The new ORF sequence, together with a downstream selection marker, was inserted exactly at the original start codon of the replaced gcs+. The homologous recombination event was confirmed by appropriate polymerase chain reaction (PCR) experiments (data not shown). This genetic engineering led to the expression of the heterologous glucosyltransferase under the control of the genuine gcs+ promoter. The glycolipid fraction of the mutant cells was analyzed by TLC showing a new glycolipid, which comigrated with authentic ßGlcß GlcD (Figure 2A). The appearance of ßGlcßGlcD in the transformed cells showed that the gcs+ promoter was active. Therefore, the gcs+ gene can be expected to be transcribed under these experimental conditions in wild type A. tumefaciens. On the other hand, in wild type cells we could not detect any glycolipids (beyond a threshold value of 0.5%) resembling in structure and extractability conventional glycolipids comprising glycosylated derivatives of DAG, ceramides, and sterols.
|
|
Expression of the Gcs from Agrobacterium tumefaciens in Pichia pastoris led to the synthesis of new glycolipids of very different structures
After these negative results concerning the identification of the enzymatic activity of the Gcs enzyme from A. tumefaciens, we switched to heterologous expression of the Gcs ORF in previously successful expression hosts. One of the hosts used before in our laboratory for functional expression of different glycosyltransferases was the yeast Pichia pastoris. This organism does not contain glycoglycerolipids, but it constitutively synthesizes ceramide glucosides and sterol glucosides as well as DAG used as intermediate for phospholipid biosynthesis. Therefore, P. pastoris represents an appropriate host for the expression of glycosyltransferases requiring DAG, sterols, or ceramides as glycosyl acceptors. We used a glycolipid-free double null mutant of P. pastoris (
gcs/ugt51B1; Hillig et al., 2003), which is devoid of both sterol and ceramide glucosyltransferase activity. This strain was transformed to express the Gcs from A. tumefaciens. The transformed cells were subjected to extraction of lipids which were fractionated and analyzed as described above.
TLC of the glycolipid fraction resolved numerous new glycolipids comigrating with glycosylceramide and various glycosylglycerolipid standards available from this and previous work (Jorasch et al., 1998; Leipelt et al., 2001) (Figure 3). A satisfactory separation of the various glycolipids by TLC required the use of different solvent mixtures. These are described in the legend of each figure and listed in Table I of the supplementary data. For structural identification glycolipids were isolated, acetylated, and subjected to compositional and structural analysis by combined gas–liquid chromatography/mass spectrometry (GLC-MS), by nuclear magnetic resonance (NMR) spectroscopy and in most cases also by electrospray ionization Fourier-transform ion cyclotron resonance mass spectrometry (ESI FT-ICR MS). By this, we isolated and identified nine different glycolipids (Figure 3). Six out of these had been described by us before (Jorasch et al., 1998, 2000; Leipelt et al., 2001; Sakaki et al., 2001) and, therefore, only data for the compounds not previously studied in our laboratory are given (Table IIa and IIb of the supplementary data). Aiming to identify the sugar specificities of the glycosyltransferases, we focussed on the glycosyl part structures, and only a few data for the aglycon lipid portions (DAG and ceramide) have been included (Table I). The expression of the Gcs from A. tumefaciens in P. pastoris induced the accumulation of six glycolipids with a DAG backbone: ßGlcD, ßGalD, ßGlcßGlcD, ßGlcßGalD, Glc-(13)-ßGlcD and ßGalßGalD (abbreviations for glycolipids used throughout the text are given in Table I).
|
The remaining three glycolipids turned out to be glycosphingolipids. Their synthesis represents at least partial support and confirmation of the sequence-based Gcs-annotation. The interpretation of the structural data resulted in the identification of two glucosyl- and one galactosylceramide (Figure 3, Table I).
The structures of the isolated glycolipids demonstrate that the expressed enzyme has an exceptionally broad substrate specificity regarding both glycosyl donor and lipophilic acceptor (Figure 4). As obvious from the different glycosyl headgroups of these new glycolipids, the Gcs of A. tumefaciens, expressed in P. pastoris, can use galactose as well as glucose donors in a promiscuous manner to form ß-linked glycosides. These glycosyl residues are transferred to a variety of lipophilic acceptors such as DAG, ceramide, ßGlcD and ßGalD, from which only DAG and ceramides are present in untransformed cells.
|
The acceptance of both DAG and glycosyldiacylglycerols results in the formation of diglycosyldiacylglycerols starting from DAG. The predominant products of these two glycosylation steps carry the terminal glycosyl residue in ß-16-linkage. The possibility that these diglycosyldiacylglycerols result from glucosylation of the corresponding monoglycosyldiacylglycerols by native glucosyltransferase activities of P. pastoris can be excluded, because expression of the human Gcs protein in P. pastoris resulted in the synthesis of ßGlcD, but not of further glycosylation products (Leipelt et al., 2001). An exception is the structure of the ßGlc(13)ßGlcD. If really attributable to the activity of the Gcs from A. tumefaciens, it would remarkably extend the relaxed specificity of this enzyme. However, at present there is no unequivocal evidence that the ßGlc(13) ßGlcD is a product of the A. tumefaciens Gcs.
Overexpression of the two bacterial Gcs sequences in Agrobacterium tumefaciens resulted in the synthesis of di- and triglycosyldiacylglycerols
Our experiments had shown a remarkably wide specificity of the Gcs sequence from A. tumefaciens when expressed in P. pastoris. But despite of this finding, Gcs function in its natural hosts is still not clear. To specify the activity of the Gcs in A. tumefaciens, the Gcs sequences were inserted into a vector for expression in A. tumefaciens. This vector was constructed by insertion of the pVS1 sequence of the origin of replication (from pCambia2200) into an Escherichia coli expression vector for stable maintenance in A. tumefaciens. Gene expression is controlled by a strong, inducible promoter. Cells of A. tumefaciens were transformed with these vectors containing in addition the ORF of the Gcs either from A. tumefaciens or from M. loti. Lipid extracts prepared from the transformed cells were analyzed by TLC, and all glycolipid components were isolated, per-O-acetylated, and analyzed by GLC-MS, ESI FT-ICR-MS, and 1H-NMR spectroscopy as described above.
The lipid patterns of the two transformants differed from each other and from the mixture found in P. pastoris (Figure 5, Table I). As expected, glycosylceramides were not present in the bacterial glycolipid fractions. Moreover, no monoglycosyldiacylglycerols could be detected, which represented a prominent fraction in the lipid extract from the transformed P. pastoris cells. On the other hand, the expression of both Gcs sequences resulted in the formation of ßGalßGalD and ßGlcßGalD, whereas ßGlcßGlcD could not be detected. The lipid extract from cells expressing the Gcs from M. loti contained additional and predominating triglycosyldiacylglycerols. The major species was identified as ßGalßGalßGalD (Table IIb of supplementary data). The band with slightly higher mobility represents a mixture which contains ßGlcßGlcßGlcD (Jorasch et al., 1998) and other triglycosyldiacylglycerols having a terminal glucose, and both galactose and glucose as sugar moieties in the two inner positions (Table I).
|
Interestingly, the two bacterial Gcs enzymes expressed in A. tumefaciens showed a preference for UDP-Gal, because lipids with only galactosyl residues (ßGalßGalD and ßGalßGalßGalD) predominated, whereas ßGlcßGlcD was not detected. These findings are in contrast to the observations made by Gcs expression in P. pastoris (Figure 3), where the glucolipids were the predominant components, although the ratio of gluco- to galactolipids seems to depend to some extent on variations in culture conditions (data not shown). The experiments on the overexpression of the Gcs sequence from A. tumefaciens under the control of a strong promoter in this bacterium suggest that the formation of ßGalßGalD and ßGlcßGalD may represent the natural activity of this Gcs in Agrobacterium.
In vitro enzymatic assays of Gcs activity support the preference for UDP-Gal
To identify the actual sugar donors and to confirm their promiscuous use by the Gcs enzyme of A. tumefaciens, we carried out in vitro enzyme assays. As enzyme source, we used membrane fractions which were prepared from both P. pastoris and E. coli cells expressing the Gcs from A. tumefaciens. The Gcs activity expressed in P. pastoris was characterized by assays with radiolabelled UDP-sugars (Figure 6A). The membrane fractions were incubated with UDP-[14C]Glc or UDP-[14C]Gal without addition of a lipophilic acceptor. After incubation, the lipids were extracted and separated by TLC followed by radioscanning. In control assays with membranes prepared from untransformed P. pastoris, no radioactivity was detected in lipids. In the assay performed with membranes from transformed cells and UDP-[14C]Glc, five components were labelled. On the basis of reference compounds, they were tentatively identified as ßGlcD, ßGalD, two different species of ßGlcCer and ßGlcßGalD (Figure 6). The appearance of ßGalD may be explained by the presence of residual epimerase activity in the membrane fraction by which part of the UDP-[14C]Glc was converted to UDP-[14C]Gal. In the corresponding assay carried out with UDP-[14C]Gal, significantly more radioactivity was incorporated into glycolipids, but a similar set of compounds was labelled. This time galactolipids were clearly dominating, whereas glucolipids were hardly labelled. On the basis of their chromatographic behaviour the major components labelled with UDP-[14C]Gal were tentatively identified as ßGalD, ßGalCer, ßGlcßGalD, and ßGalßGalD.
|
A second series of assays were performed using the membrane fraction prepared from E. coli, which were incubated with the fluorescent ceramide (D-erythro-N [6-amino-N-4' (7-nitrobenzo-2-oxa-1,3-diazolo)-hexanoyl]-ceramide) (NBD-Cer) and either unlabelled UDP-Glc or UDP-Gal (Figure 6B). After incubation, the lipophilic compounds were separated by TLC and their fluorescence was determined. Only two reaction products were detected which were tentatively identified as glucosyl- and galactosyl-NBD-Cer.
The preferred in vitro-labelling of galactolipids by the Gcs of A. tumefaciens is in line with the observation that in lipid extracts from A. tumefaciens expressing this sequence more galacto- than the corresponding glucolipids were found (Figure 5). Therefore, in vivo and in vitro data suggest that the Gcs from A. tumefaciens can use both sugar nucleotides, but UDP-Gal is significantly preferred.
For a confirmation of these conclusions, we performed kinetic measurements with the Gcs sequence from A. tumefaciens expressed in E. coli to determine the affinity of the enzyme for the two sugar nucleotides. Normally, the simplified determination of the apparent affinity for UDP-Glc and UDP-Gal is carried out in the presence of excess acceptor substrate. Because membrane preparations were used as enzyme source and neither DAG nor ceramide can be introduced into the assay mixture without the use of detergents, we used a further simplified approach by just varying the concentration of the sugar nucleotide in the presence of constant quantities of membrane protein and thus also constant, but most likely not saturating lipophilic acceptor. This approach resulted in apparent KM- and Vmax- values which can only be used for comparative purposes. The membrane fraction was prepared from cells of E. coli C41(DE3) expressing the Gcs from A. tumefaciens. It was used for in vitro enzyme assays in the presence of different concentrations of either UDP-[14C]Gal or UDP-[14C]Glc. First, analysis of the reaction products by TLC showed that glycolipids were formed (data not shown), which were identical to the products of the assay performed with A. tumefaciens cells overexpressing the Gcs sequence (Figure 5). Subsequently, the incorporated total radioactivity from additional assays was measured by scintillation counting. In the experiments with UDP-Gal, sufficient radioactivity was incorporated into lipids, whereas in the assays with UDP-Glc, performed under the same conditions, the measured values were too low to be used for a reliable evaluation. With UDP-Gal typical Michaelis-Menten-like kinetics were observed (data not shown). Substrate concentrations at Vmax/2 between 40 and 60 µM UDP-Gal were calculated with Vmax-values ranging between 30 and 40 pmol/min/mg. These experiments show that the Gcs enzyme expressed in E. coli has a significantly higher affinity for UDP-Gal.
|
Discussion |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsSupplementary dataAcknowledgmentsReferences |
|---|
The Gcs enzymes from A. tumefaciens and M. loti are the first bacterial representatives of the GT21 family which have been characterized experimentally. Despite a significant sequence similarity to other members in GT21, both Gcs enzymes from A. tumefaciens and M. loti differ from known GlcCer synthases in three features. These are sugar acceptor specificity, sugar donor specificity, and "processivity," which will be discussed in the following. The different glycosylation alternatives and the resulting products are summarized in Figure 4.
Gcs enzymes from animals, fungi, and plants transfer a sugar moiety to ceramide, whereas the Gcs enzymes from the two bacteria use DAG as acceptor molecule. This finding may be explained by the fact that DAG and ceramide are structurally similar (Jorasch et al., 2000), that A. tumefaciens and M. loti apparently do not contain ceramides and that broad sugar acceptor specificity regarding DAG and ceramide is a common feature of many other glycosyltransferases (Jorasch et al., 1998, 2000). The overexpression of the human and the bacterial Gcs enzymes in P. pastoris, which contains both DAG and ceramide, revealed this broad specificity concerning the glycosyl acceptor. While expression of the human Gcs resulted in the glycosylation of mainly ceramides with lower proportions of glycosylated DAG (Leipelt et al., 2001), we here could demonstrate that the Gcs from A. tumefaciens synthesized predominantly glycosyldiacylglycerols and lower proportions of glycosylceramides. A further erosion of acceptor specificity is typical for another Gcs, cloned from cotton, which on overexpression produces glycosylceramides and sterol glycosides (Hillig et al., 2003). In conclusion, many Gcs enzymes from different species exhibit broad acceptor specificity, but the Gcs from A. tumefaciens and M. loti differ from Gcs enzymes of animals, fungi, and plants by their preference of DAG over ceramide (Table II).
The second differing feature is the sugar-donor specificity. While all eukaryotic Gcs use UDP-Glc as sugar donor, the two bacterial glycosyltransferases favour UDP-Gal over UDP-Glc (Table II). In vitro enzyme assays have shown that the Km of mammalian Gcs for UDP-Glc is at least 200 times lower than for UDP-Gal (Sprong et al., 1998; Wu et al., 1999). Because eukaryotic Gcs activity always results in the synthesis of GlcCer but not of galactosylceramide in vivo, these enzymes are referred to as glucosyltransferases (Ichikawa et al., 1996; Sprong et al., 1998; Wu et al., 1999; Leipelt et al., 2001). Although we were not able to determine the Km of the bacterial Gcs for UDP-Glc, we demonstrated by in vitro assays that the enzymes transfer both galactosyl and glucosyl moieties with a pronounced preference for galactose. These in vitro data are reflected by the isolation of both galactosylated and glucosylated products from hosts expressing the bacterial Gcs enzymes. The synthesis of, for example, ßGlcßGalD by the A. tumefaciens Gcs in its natural host justifies that this Gcs is referred to as a galactosyl/glucosyltransferase.
Relaxed substrate specificity of an enzyme concerning donor and acceptor is an interesting phenomenon in terms of the reaction mechanism and the structure of its reaction centre which is not the subject of our present work. In addition, this characteristic may have consequences for the biological functions of the enzyme. For most prokaryotic organisms, inaccurate recognition of the sugar acceptor by the Gcs enzyme may not be relevant, because they contain only DAG. In contrast, eukaryotic organisms contain both DAG and ceramide, which not only serve as backbone for membrane lipid biosynthesis, but each also being involved in different signalling cascades. Therefore, eukaryotes would be expected to have Gcs enzymes with strict acceptor specificity, or there should be precautions for a spatial separation between the enzyme and "unwanted" substrates. This assumption is generally applicable to other enzymes acting on DAG and ceramide. Because many of these eukaryotic enzymes apparently do not show strict specificity (Jorasch et al., 1998; Leipelt et al., 2001; Tadano-Aritomi and Ishizuka, 2003), it seems likely that in most of the cases the enzymes are specifically targeted to intracellular membrane systems to confine their contact to the "desired" substrates.
In contrast to relaxed specificity for the sugar acceptor, most glycosyltransferases show a strict specificity for the sugar donor. Eukaryotes, for example, use two different enzymes to synthesize glucosyl- and galactosylceramide, respectively (Schulte and Stoffel, 1993; Ichikawa et al., 1996). Thus, the transfer of alternative sugar moieties to the acceptor, in particular the transfer of glucosyl or galactosyl residues by bacterial Gcs enzymes as demonstrated by our data, is one of the few exceptions to this rule. Another exception is the 
-N-acetylhexosaminyltransferase of animal tissues catalyzing the transfer of both N-acetylglucosamine and N-acetylgalactosamine (Lind et al., 1998; Pedersen et al., 2003). An even more interesting representative of these exceptions is the glycosyltransferase LpsB of Sinorhizobium meliloti—whose natural activity may be the transfer of glucose—with the ability to complement a mutant of Rhizobium leguminosarum defective in the orthologous and highly selective mannosyltransferase LpcC (Kanipes et al., 2003). In this case, the unspecificity in substrate acceptance does not only involve two epimeric sugar moieties (glucose and mannose), but most likely also different nucleotide portions (UDP and GDP).
It is generally assumed that given glycosyl moieties play specific roles in glycolipids, glycoproteins and polysaccharides which cannot be fulfilled by different sugar moieties. Blood group factors are one of the many examples which confirm this assumption. However, concerning simple glycosphingolipids and other glycolipids there are only very few data regarding the specific functions of particular sugar moieties. It is, for example, still unclear whether the functions of common glycolipids such as ß-GlcCer, sterol-ß-glucoside, ß-galactosylceramide, ß-galactosyldiacylglycerol, and -galactosyl-(16)-ß-galactosyldiacylglycerol could be fulfilled by corresponding lipids with a different glycosyl moiety or with an identical sugar moiety of different anomeric structure. This in turn would provide evolutionary pressure on maintaining or widening substrate specificities including the chance to develop new characteristics.
The third differing feature is "processivity." In this context, "processivity" means the successive transfer of glycosyl residues to a lipid acceptor in the first step and to glycolipids with a growing glycan chain in the following steps. We could however not demonstrate that the successive addition of sugar residues occurred without dissociation of the enzyme-acceptor complex. In all cases which lack evidence for such a mechanism we therefore marked the term "processive" by inverted commas. While eukaryotic members of GT21 exclusively synthesize monoglycosylated lipids, the bacterial Gcs successively transfer one or two additional sugar moieties in ß16-linkage to the monoglycosyldiacylglycerol. In agreement with the "processivity" of bacterial Gcs enzymes, their sequences contain a D1,D2,D3(Q/R)XXRW motif, which was previously identified as a characteristic feature of processive ß-glycosyltransferases of GT2 such as cellulose synthase or chitin synthase (Saxena et al., 1995). Interestingly, the nonprocessive eukaryotic members of GT21 exhibit a more or less complete D1,D2,D3(Q/R)XXRW motif as well (Table II) (Leipelt et al., 2001), which is essential for enzymatic activity (Marks et al., 2001). Therefore, it can be assumed that this motif is required but not sufficient for "processivity" of the bacterial GT21 glycosyltransferases.
"Processivity" is a common feature of other lipid glycosyltransferases, for example from Bacillus subtilis and S. aureus. Their GT28 glucosyltransferases (YpfP) synthesize ßGlcD, ßGlcßGlcD, and ßGlcßGlcßGlcD which serve as membrane lipids and glycolipid anchors of lipoteichoic acids (Jorasch et al., 1998, 2000; Kiriukhin et al., 2001). Plants contain a hypothetical "processive" galactosyltransferase activity which may form the series of polygalactolipids up to a pentagalactosyldiacylglycerol (Heinz, 1996; Kelly et al., 2003; Xu et al., 2003). Several higher glycosylated derivatives of ceramide, DAG, and sterols occur in plants and fungi, which may suggest the existence of additional, but so far unknown "processive" lipid glycosyltransferases (Heinz, 1996; Sperling et al., 2004). Despite the ubiquitous occurrence of products of "processive" lipid glycosyltransferases, the function of their products are mainly unknown except for the hypothesis that sitosterol cellodextrins serve as precursors for cellulose biosynthesis in plants (Peng et al., 2002).
In this work we have characterized in vivo and in vitro activities of the bacterial Gcs, but the biological functions of these glycosyltransferases in A. tumefaciens and M. loti are still not clear. Deletion of the gcs+ gene in A. tumefaciens was not lethal, and the mutant cells grew like the wild type under the given laboratory conditions. Although the gcs+ promoter was active, the corresponding glycosyltransferase products of wild type cells could not been detected. Therefore, we conclude that only trace amounts of Gcs products are synthesized in A. tumefaciens, which are below the limits of our detection system. Such proportions, however, may be sufficient to fulfil biological functions, for example in signalling chains. Another possibility is, that expression of the gcs+ gene is very low only under laboratory conditions, but may increase under changed conditions (e.g., stress, plant-infection). This could lead to the synthesis of considerable amounts of glycolipids which may fulfil their function as membrane components or serve as precursors, for example, cell wall polymers.
In this context it should be mentioned that some Rhizobiaceae and related species change their lipid composition under low oxygen stress which occurs in culture or in symbiosis with plants. For example, Bradyrhizobium accumulates phosphatidyl inositol (Tang and Hollingsworth, 1998), whereas S. meliloti and Rhodobacter sphaeroides respond to phosphate limiting conditions by the synthesis of phosphate-free membrane lipids (Benning et al., 1995; Geiger et al., 1999). During plant–microbe interaction, R. leguminosarum accumulates diglycosyldiacylglycerol which in turn elicits various symbiosis-relevant morphological changes of the host plant (Orgambide et al., 1994). In view of these data, the identification of the Gcs activities of A. tumefaciens and M. loti provides a basis for further studies on glycolipid functions in parasitic or symbiontic bacteria.
|
Materials and methods |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsSupplementary dataAcknowledgmentsReferences |
|---|
Bacterial and yeast strains, growth and recombinant DNA techniques
Escherichia coli strains XL1-Blue (MRF') (Stratagene, La Jolla, CA) and C41(DE3) (Miroux and Walker, 1996)
were routinely grown aerobically at 37°C in Luria-Bertani medium (Sambrook et al., 1989). Ampicillin (100 mg·L–1), kanamycin (30 mg·L–1), or chloramphenicol (30 mg·L–1) were included for growth of plasmid-bearing cells. Agrobacterium tumefaciens strain ATHVC58C1, a derivative from the strain EHA 101 (Hood et al., 1986), which was kindly provided by Dr. J. Dettendorfer (KWS, Einbeck, Germany) was grown at 28°C in YEP (10 g·L–1 peptone, 10 g·L–1 yeast extract, 5 g·L–1 NaCl, pH 7.2) in the presence of rifampicin (80 mg·L–1). Kanamycin (50 mg·L–1) or chloramphenicol (50 mg·L–1) was included for growth of plasmid-bearing cells. Streptomycin (300 mg·L–1) and spectinomycin (100 mg·L–1) or kanamycin (50 mg·L–1) were included to select A. tumefaciens after homologous recombination. The yeast strain used in this study was P. pastoris JC 308 gcs/ugt51B1 (Hillig et al., 2003), grown at 30°C in YPG medium (10 g·L–1 yeast extract, 20 g·L–1 peptone, 10 mL×L–1 glycerol). For gene expression driven by the AOX1 promoter, 0.5% methanol was added to the growth medium. The vectors pET24d(+) (Novagen, Madison, WI), pUC18 (Yanish-Perron et al., 1989), pBluescript (Stratagene), pTrcHis2 C (Invitrogen, Carlsbad, CA), pCambia2200 (Cambia, Canberra, Australia), pLH7000 (Hausmann and Toepfer, 1999), pFP1-3 (Götz et al., 1999), pEsay24 (Jorasch et al., 2000), and pPIC3.5 (Invitrogen) were used for cloning. Standard methods were followed for DNA isolation, restriction endonuclease analysis, and ligation (Sambrook et al., 1989).
Cloning of the Gcs from Agrobacterium tumefaciens and Mesorhizobium loti
The ORF sequences of the gcs+ genes from A. tumefaciens and M. loti were amplified from genomic DNA by PCR using the specific oligonucleotide primer pairs 1F/1R or 2F/2R, respectively (Table III of supplementary data which also lists all other primers). Pfu polymerase (Stratagene) was used for the amplification of the 1164 bp and 1152 bp products containing the entire Gcs ORF sequences of A. tumefaciens and M. loti. These amplicons were inserted into a SmaI-linearized pUC18 vector resulting in pUCAGRO and pUCMESO which were used for transformation of E. coli XL1Blue cells. Exact in-frame cloning and identity of the PCR-cloned fragments were confirmed by sequencing. The PCR fragment corresponding to the gcs+ gene of A. tumefaciens was also cloned into pPIC3.5/SnaBI resulting in pPICAGRO to be used as expression vector in P. pastoris JC 308 gcs/ugt51B1.
For expression in E. coli C41(DE3) the A. tumefaciens gcs+ ORF sequence was amplified with the primers 3F and 3R using pUCAGRO as template. After subcloning into pUC18 the insert was released with NcoI/BamHI and ligated with the NcoI/BamHI-linearised expression vector pET24d.
To provide the Gcs nucleotide sequences of A. tumefaciens and M. loti with the restriction sites AvrII and BamHI, they were amplified with the specific oligonucleotide primers 8F and 8R using pUCAGRO as template and 9F and 9R with pUCMESO as template. The purified PCR products were inserted into the SmaI-linearized pUC18 vector resulting in p18agro and p18meso. The inserts from the AvrII/BamHI-digested vectors p18agro and p18meso were ligated with the AvrII/BamHI-linearised expression vector pTnsyn3, resulting in pTnagro and pTnmeso. pTnsyn3, pTnagro and pTnmeso were originally used for the expression of different glycosyltransferases in cyanobacteria (unpublished data), but in this study pTnagro and pTnmeso were used for the Gcs expression in E. coli XL1-Blue. pTnsyn3 was created by the ligation of pTrcHis2 C (NcoI/HindIII) with the inserts from the vectors p18Nsyn3 released with NcoI/BamHI and p18BnptIIH released with BamHI/HindIII. p18Nsyn3 resulted from the ligation of pUC18 (SmaI) with the PCR product of a synthetic sequence that provided the restriction sites NcoI, AvrII and BamHI, the corresponding primers were 11F and 11R. p18BnptIIH bears a kanamycin resistance (KanR)-cassette which was amplified from pFP1-3 (Götz et al., 1999) with the primers 10F and 10R.
pTnVagro and pTnVmeso were used as expression vectors in A. tumefaciens. These vectors resulted from pTnagro and pTnmeso linearised with BstZ17I and ligated with the pVS1 rep sequence from pCambia2200 (BstZ17I/HincII).
Deletion of the genomic gcs+ sequence from Agrobacterium tumefaciens and expression of a heterologous glucosyltransferase in this mutant
The vector pBa1na5 was used to disrupt the gcs+ gene of A. tumefaciens by homologous recombination. The disruption was performed by replacement of the 5'-end of the gcs+ gene sequence comprising about 385 bp plus the flanking sequence of about 700 bp upstream of the gcs+ ORF sequence by a KanR -cassette (Figure 2C). The vector pBa1na5 is the ligation product of the vector pBa1n (EcoRV/SpeI) and the 5'-truncated ORF of the gcs+ from A. tumefaciens (blunted at 5'-end and SpeI-digested at 3'-end). This fragment was generated by the digestion of the Gcs ORF in p18agro with AscI to cut off the 5'-end followed by blunting this restriction site by a fill-in reaction. The fragment was released with SpeI. pBa1 is the ligation product of pBluescript (KpnI/SalI) with two inserts which were the 3'-truncated flanking sequence released from p18agro1 (KpnI/HindIII) and the KanR-cassette from p18BnptIIH (HindIII/SalI).
The vector pBa1saySSa2 was used to disrupt the gcs+ gene from A. tumefaciens with simultaneous insertion of the heterologous glucosyltransferase from S. aureus (ugt106B1) behind the native gcs+ promoter of A. tumefaciens (Figure 2D). For this purpose, about 600 bp at the 5'-end of the gcs+ gene sequence were replaced by the sequence of the heterologous glucosyltransferase from S. aureus together with a streptomycin/spectinomycin resistance (SmR/SpR)-cassette. The S. aureus glucosyltransferase was inserted exactly at the locus of the replaced gcs+ ORF.
The vector pBa1saySSa2 was constructed in several steps with successive integration of four fragments beginning with the linearised vector pBluescript (KpnI/EcoRV). These fragments were released from p18agro1 (KpnI/EcoRV; left flanking sequence), p18Staph (EcoRV/SmaI; heterologous glycosyltransferase), p18SS (SmaI/NotI; SmR/SpR-cassette), and p18agro2 (NotI/SacI; right flanking sequence). The vector p18agro1 resulted from the ligation of pUC18 (SmaI) with the 2100 bp flanking sequence upstream of the gcs+ gene of A. tumefaciens which was amplified from genomic DNA with the primers 4F and 4R. A 600 bp sequence corresponding to the 3'-end of the gcs+ gene of A. tumefaciens was amplified with the primers 5F and 5R and ligated with pUC18 (SmaI) resulting in the vector p18agro2. The "processive" glucosyltransferase from S. aureus was amplified with the primers 6F and 6R using the vector pEsay24 as template and ligated with pUC18 (SmaI) resulting in the vector p18Staph. The SmR/SpR-cassette was amplified from pLH7000 with the primers 7F and 7R and inserted into pUC18 (SmaI) giving p18SS. For transformation, the vectors pBa1na5 and pBa1saySSa2 were digested with KpnI/SacI to release the linearised transformation constructs. They were used for transformation of competent A. tumefaciens cells by electroporation. Kanamycin- or streptomycin/spectinomycin-resistant transformants were selected by growth on YEP plates containing the appropriate antibiotics.
Replacement of the wild type gcs+ gene sequence was monitored by PCR with Taq DNA Polymerase (NEB) or HerculaseR Enhanced DNA Polymerase (Stratagene) (data not shown). To check the deletion of the gcs+ gene and its replacement by the KanR-cassette the following primer pairs were used (Figure 2B–D, Table III): 12F/16R—to check the correct insertion at the 5'-end; 15F/15R—to check the correct insertion at the 3'-end; 12F/15R—to check the complete replacement comprising the region from the 5'- to 3'-end. The replacement by the ugt106B1 sequence from S. aureus plus the SmR/SpR-cassette was examined by the following primer pairs: 12F/12R (5'-end); 13R/13F (3'-end); 14F/14R (complete replacement, 5'- to 3'-end).
Heterologous expression of the Gcs ORF sequences from Agrobacterium tumefaciens and Mesorhizobium loti in different hosts
Escherichia coli XL1-Blue (MRF') was used as expression host for the vectors pTrcHis2 C, pTnagro and pTnmeso, whereas E. coli C41(DE3) was used as expression host for the vectors pET24d and pETagro. Cultures of E. coli were grown overnight at 37°C. Expression cultures were started at OD600 = 0.05 and grown to an optical density of 0.8. Induction was performed by adding 0.4 mM isopropyl thio-ß-D-galactoside and further incubation for 4 h at 37°C. Cells were collected by centrifugation (15 min, 5000 g). Pichia pastoris JC 308 gcs/ugt51B1 cells were grown at 30°C in YPG medium to an OD600 between 1 and 2. Expression was driven by the strong AOX1 promoter and induced by addition of 0.5% methanol to the medium. Cells were harvested by centrifugation (10 min, 5000 g) 20 h after induction. A. tumefaciens cells were used as expression host for the vectors pTnVagro and pTnVmeso. The cultures were grown at 28°C to an OD600 = 1.0 and induced by adding 0.4 mM isopropyl thio-ß-D-galactoside and further incubation for 15 h at 28°C. Cultures of A. tumefaciens wild type and knock out mutants, including mutants containing the gene of the heterologous glucosyltransferase of S. aureus, were grown at 28°C to an OD600 = 2.0 without induction. The A. tumefaciens cells were harvested by centrifugation (30 min, 8000 g).
Enzymatic assays of glycosyltransferase activities
For enzymatic assays, the cell pellets recovered from 25 mL of the E. coli, A. tumefaciens, or P. pastoris expression cultures were resuspended in 0.5 mL of buffer 1 (50 mM Tris–HCl, pH 8.0; 7%, v/v, glycerol). All subsequent steps were carried out at 4°C. The E. coli or A. tumefaciens cells were disrupted by ultrasonication (eight times for 10 s). The P. pastoris cells were disrupted by adding 2–3 g of glass beads (0.4 mm diameter) and vortexing for 30 s followed by cooling on ice for 30 s. This procedure was repeated 10 times. Cell debris were removed by centrifugation (1 min, 2800 g). Supernatant fractions were centrifuged at 100,000 g for 30 min, and the sedimented membrane fraction was resuspended in 400 µL (E. coli or A. tumefaciens) or 800 µL (P. pastoris) of buffer 2 (50 mM Tris–HCl, pH 7.6; 7%, v/v, glycerol).
The actual assays were performed in a final volume of 100 µL of buffer 2. Radioactive assays were supplied with UDP-[14C]Glc (150,000 dpm, specific activity 10 GBq/mmol, final concentration 2.5 µM) or UDP-[14C]Gal (150,000 dpm, specific activity 10.9 GBq/mmol, final concentration 2.3 µM). Assays with NBD-Cer (Matreya, Pleasant Gap, PA; final concentration 0.01 µg/µL), were supplied with unlabelled UDP-Glc or UDP-Gal (each in a final concentration of 500 µM). Each assay was supplied with 50 µL of the resuspended membrane fraction and incubated for 90 min at 30°C. The reaction was terminated by the addition of 3 mL chloroform/methanol (2:1, v/v) and 0.75 mL of 0.45% NaCl solution (w/v). The extracted lipids were separated by TLC. Radioactivity on TLC plates was detected by radioscanning with a BAS-1000 BioImaging Analyzer (Raytest, Straubenhardt, Germany). NBD-Cer fluorescence on TLC plates was scanned using an AlphaDigiDocTM Gel Documentation & Image Analysis System (Alpha Innotech Corporation, San Leandro, CA).
Lipid extraction and analysis
Expression cultures of E. coli, P. pastoris, and A. tumefaciens were grown and harvested as described above. The sedimented cells were boiled for 10 min in water and centrifuged again. Lipid extraction was performed with chloroform/methanol 1:2 (v/v) and chloroform/methanol 2:1 (v/v). The lipid extract was washed by Folch partitioning (Folch et al., 1957) (CHCl3:methanol:0.45% NaCl solution, 2:1:0.75), and the organic phase was evaporated. The residue was redissolved in CHCl3 and fractionated by chromatography on SPE SI-1 columns (Phenomenex, Torrance, CA). Neutral lipids were eluted with chloroform, the glycolipid fraction was obtained with acetone/isopropanol 9:1 (v/v), and phospholipids were eluted with methanol. These fractions were subjected to analytical and preparative TLC separations using CAMAG Automatic TLC Sampler 4 (CAMAG, Muttenz, Schweiz) for spotting. Glycolipids were visualized by spraying with -naphthol/sulphuric acid and subsequent heating to 160°C. Preparative separations were performed by TLC on silica gel 60 (Merck, Darmstadt, Germany). The solvents used to separate the different glycolipids are given in Table I of supplementary data. Lipids were visualized under UV light after spraying with ANS solution (8-anilino-1-naphthalenesulfonic acid ammonium salt, 0.2%, w/v in methanol). For NMR spectroscopy and mass spectrometry (MS), the purified glycolipids were acetylated (with acetic anhydride in pyridine, 1:1) overnight at room temperature and subjected to repurification by preparative TLC in diethyl ether.
Structural analysis
Compositional analysis. Fatty acids in the glycolipids were methanolyzed with 0.5 M HCl in methanol at 85°C for 45 min. After removal of the solvent, the products were peracetylated with acetic anhydride in pyridine (1:1.5, v/v, 85°C, 20 min). For analysis of the sugar components glycolipids were methanolyzed under stronger conditions (2 M HCl in methanol at 85°C for 16 h), then hydrolyzed with aqueous 4 M CF3CO2H at 100°C for 2 h, conventionally reduced with borohydride and peracetylated (Sawardeker et al., 1965).
Methylation analysis. Purified glycolipids were methylated with CH3I in dimethyl sulfoxide in the presence of solid NaOH (Ciucanu and Kerek, 1984) and subsequently hydrolyzed with 2 M CF3CO2H (100°C, 2 h). The partially methylated monosaccharides were reduced with borohydride, peracetylated, and analyzed by GLC and GLC-MS as described below.
GLC-MS analysis. The sugar and fatty acid derivatives were analyzed by GLC on a Hewlett-Packard HP 5890 Series II chromatograph, equipped with a 30-m fused-silica SPB-5 column (Supelco, St. Louis, MO) using a temperature gradient of 150°C (3 min) 320°C at 5°/min, and GLC-MS on a Hewlett-Packard HP 5989A instrument equipped with a 30-m HP-5MS column (Hewlett-Packard, Palo Alto, CA) and operated under the same conditions (Zähringer et al., 1997).
ESI FT-ICR MS. High resolution FT-MS was performed in the positive ion modes using an APEX II–instrument (Bruker Daltonics, Billerica, MA) equipped with an actively shielded 7 Tesla magnet and an (nano) ESI ion source. Mass spectra were acquired using standard experimental sequences as provided by the manufacturer. Samples were dissolved in a 50:50:0.03 (v/v/v) mixture of 2-propanol, water, 30 mM ammonium acetate adjusted with acetic acid to pH 4.5 at a concentration of 10 ng.µL–1. The samples were sprayed at a flow rate of 2 µL.min–1, and the drying gas temperature was set to 150°C. Each spectrum represents an average of at least 20 transients composed of 1 M data points.
Proton (1H) nuclear magnetic resonance spectroscopy. All one- (1D) and two- (2D) dimensional homonuclear 1H-NMR spectra were recorded at 600 MHz (Bruker Avance DRX 600, Bruker Instruments, Billerica, MA). Before the measurements, the purified per-O-acetylated glycolipids (25–500 µg) were exchanged twice from 2HCCl3 (99.8%D, Aldrich, St. Louis, MO). All 1D and 2D 1H-NMR spectra (COSY, RELAY, TOCSY) were recorded in 3 mm microtubes (Kontes, Vineland, NJ) in 2HCCl3 (99.96%, Eurisotop, Gif-sur-Yvette, Saint-Aubin France) at 300 K. Chemical shifts were referenced to internal chloroform (
H = 7.260 ppm). Bruker standard software XWINNMR 3.5 was used to acquire and process all 1D and 2D spectra. A selection of analytically relevant data of new compounds is given in Table II of supplementary data.
|
Supplementary data |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsSupplementary dataAcknowledgmentsReferences |
|---|
Supplementary data are available at Glycobiology online (http://glycob.oupjournals.org/).
|
Acknowledgments |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsSupplementary dataAcknowledgmentsReferences |
|---|
We thank H. Moll, U. Schombel and B. Kunz, Research Center Borstel, and W. Hellmeyer, Biozentrum Klein Flottbek for excellent technical assistance. This study was supported by the DFG Sonderforschungsbereich SFB 470.
|
Abbreviations |
|---|
DAG, diacylglycerol(s); ESI FT-ICR MS, electrospray ionization Fourier-transform ion cyclotron resonance mass spectrometry; Gcs, glucosylceramide synthase; GLC-MS, combined gas–liquid chromatography/mass spectrometry analysis; GT21, glycosyltransferase family 21; MS, mass spectrometry; NBD-Cer, D-erythro-N [6-amino-N-4'(7-nitrobenzo-2-oxa-1,3-diazolo)-hexanoyl]-ceramide; NMR, nuclear magnetic resonance; ORF, open reading frame; PCR, polymerase chain reaction; TGD, triglycosyl diacylglycerol; THF, tetrahydrofuran; TLC, thin layer chromatography; UDP-Gal, uridine diphosphogalactose; UDP-Glc, uridine diphosphoglucose
|
References |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsSupplementary dataAcknowledgmentsReferences |
|---|
Altschul, S.F., Gish, W., Miller, W., Myers, E.W., and Lipman, D.J. (1990) Basic local alignment search tool. J. Mol. Biol., 215, 403–410.[CrossRef][Web of Science][Medline]
Benning, C., Huang, Z.H., and Gage, D.A. (1995) Accumulation of a novel glycolipid and a betaine lipid in cells of Rhodobacter sphaeroides grown under phosphate limitation. Arch. Biochem. Biophys., 317, 103–111.[CrossRef][Web of Science][Medline]
Campbell, J.A., Davies, G.J., Bulone, V., and Henrissat, B. (1997) A classification of nucleotide-diphospho-sugar glycosyltransferases based on amino acid sequence similarities. Biochem. J., 326, 929–939.[Web of Science][Medline]
Ciucanu, I. and Kerek, F. (1984) A simple and rapid method for the permethylation of carbohydrates. Carbohydr. Res., 131, 209–217.[CrossRef][Web of Science]
Coutinho, P.M., Deleury, E., Davies, G.J., and Henrissat, B. (2003) An evolving hierarchical family classification for glycosyltransferases. J. Mol. Biol., 328, 307–317.[CrossRef][Web of Science][Medline]
Folch, J., Lees, M., and Sloane Stanley, G.H. (1957) A simple method for the isolation and purification of total lipides from animal tissues. J. Biol. Chem., 226, 497–509.
Futerman, A.H. and Pagano, R.E. (1991) Determination of the intracellular sites and topology of glucosylceramide synthesis in rat liver. Biochem. J., 280, 295–302.[Medline]
Geiger, O., Rohrs, V., Weissenmayer, B., Finan, T.M., and Thomas-Oates, J.E. (1999) The regulator gene phoB mediates phosphate stress-controlled synthesis of the membrane lipid diacylglyceryl-N, N, N-trimethylhomoserine in Rhizobium (Sinorhizobium) meliloti. Mol. Microbiol., 32, 63–73.[CrossRef][Web of Science][Medline]
Götz, T., Windhövel, U., Boger, P., and Sandmann, G. (1999) Protection of photosynthesis against ultraviolet-B radiation by carotenoids in transformants of the cyanobacterium Synechococcus PCC7942. Plant Physiol., 120, 599–604.
Hausmann, L. and Toepfer, R. (1999) Entwicklung von Plasmidvektoren. In Toepfer, R. (ed.), Vorträge für Pflanzenzüchtung. Bioengineering für Rapssorten nach Maß. Göttingen, Vol. 45, pp. 155–172.
Heinz, E. (1996) Plant glycolipids: structure, isolation and analysis. In Christie, W.W. (ed.), Advances in Lipid Methodology—Three. The Oily Press LTD, Dundee, Vol. 7, pp. 211–332.
Hillig, I., Leipelt, M., Ott, C., Zähringer, U., Warnecke, D., and Heinz, E. (2003) Formation of glucosylceramide and sterol glucoside by a UDP-glucose-dependent glucosylceramide synthase from cotton expressed in Pichia pastoris. FEBS Lett., 553, 365–369.[CrossRef][Web of Science][Medline]
Hood, E.E., Helmer, G.L., Fraley, R.T., and Chilton, M.D. (1986) The hypervirulence of Agrobacterium tumefaciens A281 is encoded in a region of pTiBo542 outside of T-DNA. J. Bacteriol., 168, 1291–1301.
Ichikawa, S. and Hirabayashi, Y. (1998) Glucosylceramide synthase and glycosphingolipid synthesis. Trends Cell Biol., 8, 198–202.[CrossRef][Web of Science][Medline]
Ichikawa, S., Sakiyama, H., Suzuki, G., Hidari, K.I., and Hirabayashi, Y. (1996) Expression cloning of a cDNA for human ceramide glucosyltransferase that catalyzes the first glycosylation step of glycosphingolipid synthesis. Proc. Natl. Acad. Sci. U. S. A., 93, 12654.
Jorasch, P., Wolter, F.P., Zähringer, U., and Heinz, E. (1998) A UDP glucosyltransferase from Bacillus subtilis successively transfers up to four glucose residues to 1,2-diacylglycerol: expression of ypfP in Escherichia coli and structural analysis of its reaction products. Mol. Microbiol., 29, 419–430.[CrossRef][Web of Science][Medline]
Jorasch, P., Warnecke, D.C., Lindner, B., Zähringer, U., and Heinz, E. (2000) Novel processive and nonprocessive glycosyltransferases from Staphylococcus aureus and Arabidopsis thaliana synthesize glycoglycerolipids, glycophospholipids, glycosphingolipids and glycosylsterols. Eur. J. Biochem., 267, 3770–3783.[Web of Science][Medline]
Kanipes, M.I., Kalb, S.R., Cotter, R.J., Hozbor, D.F., Lagares, A., and Raetz, C.R. (2003) Relaxed sugar donor selectivity of a Sinorhizobium meliloti ortholog of the Rhizobium leguminosarum mannosyl transferase LpcC. Role of the lipopolysaccharide core in symbiosis of Rhizobiaceae with plants. J. Biol. Chem., 278, 16365–16371.
Kawahara, K., Seydel, U., Matsuura, M., Danbara, H., Rietschel, E.T., and Zähringer, U. (1991) Chemical structure of glycosphingolipids isolated from Sphingomonas paucimobilis. FEBS Lett., 292, 107–110.[CrossRef][Medline]
Kawahara, K., Moll, H., Knirel, Y.A., Seydel, U., and Zähringer, U. (2000) Structural analysis of two glycosphingolipids from a lipopolysaccharide lacking bacterium Sphingomonas capsulata. Eur. J. Biochem., 267, 1837–1846.[Web of Science][Medline]
Kelly, A.A., Froehlich, J.E., and Dormann, P. (2003) Disruption of the two digalactosyldiacylglycerol synthase genes DGD1 and DGD2 in Arabidopsis reveals the existence of an additional enzyme of galactolipid synthesis. Plant Cell, 15, 2694–2706.
Kiriukhin, M.Y., Debabov, D.V., Shinabarger, D.L., and Neuhaus, F.C. (2001) Biosynthesis of the glycolipid anchor in lipoteichoic acid of Staphylococcus aureus RN4220: role of YpfP, the diglucosyldiacylglycerol synthase. J. Bacteriol., 183, 3506–3514.
Kohyama-Koganeya, A., Sasamura, T., Oshima, E., Suzuki, E., Nishihara, S., Ueda, R., and Hirabayashi, Y. (2004) Drosophila glucosylceramide synthase: a negative regulator of cell death mediated by proapoptotic factors. J. Biol. Chem., 279, 35995–36002.
Kojima, M., Seki, K., Ohnishi, M., Ito, S., and Fujino, Y. (1990) Structure of novel glyceroglycolipids in Adzuki bean (Vigna angularis) seeds. Biochem. Cell Biol., 68, 59–64.[Web of Science][Medline]
Lee, L., Abe, A., and Shayman, J. (1999) Improved inhibitors of glucosylceramide synthase. J. Biol. Chem., 274, 14662–14669.
Leipelt, M., Warnecke, D., Zähringer, U., Ott, C., Müller, F., Hube, B., and Heinz, E. (2001) Glucosylceramide synthases, a gene family responsible for the biosynthesis of glucosphingolipids in animals, plants, and fungi. J. Biol. Chem., 276, 33621–33629.
Lind, T., Tufaro, F., McCormick, C., Lindahl, U., and Lidholt, K. (1998) The putative tumor suppressors EXT1 and EXT2 are glycosyltransferases required for the biosynthesis of heparan sulfate. J. Biol. Chem., 273, 26265–26268.
Marks, D.L., Wu, K., Paul, P., Kamisaka, Y., Watanabe, R., and Pagano, R.E. (1999) Oligomerization and topology of the Golgi membrane protein glucosylceramide synthase. J. Biol. Chem., 274, 451–456.
Marks, D.L., Dominguez, M., Wu, K., and Pagano, R.E. (2001) Identification of active site residues in glucosylceramide synthase. A nucleotide-binding catalytic motif conserved with processive beta-glycosyltransferases. J. Biol. Chem., 276, 26492–26498.
Miroux, B. and Walker, J.E. (1996) Over-production of proteins in Escherichia coli: mutant hosts that allow synthesis of some membrane proteins and globular proteins at high levels. J. Mol. Biol., 260, 289–298.[CrossRef][Web of Science][Medline]
Olsen, I. and Jantzen, E. (2001) Sphingolipids in bacteria and fungi. Anaerobe, 7, 103–112.[CrossRef][Web of Science]
Orgambide, G.G., Hollingsworth, R.I., and Dazzo, F.B. (1992) Structural characterization of a novel diglycosyl diacylglyceride glycolipid from Rhizobium trifolii ANU843. Carbohydr. Res., 223, 151–159.
Orgambide, G.G., Philip-Hollingsworth, S., Hollingsworth, R.I., and Dazzo, F.B. (1994) Flavone-enhanced accumulation and symbiosis-related biological activity of a diglycosyl diacylglycerol membrane glycolipid from Rhizobium leguminosarum biovar trifolii. J. Bacteriol., 176, 4338–4347.
Pedersen, L.C., Dong, J., Taniguchi, F., Kitagawa, H., Krahn, J.M., Pedersen, L.G., Sugahara, K., and Negishi, M. (2003) Crystal Structure of an 1,4-N-Acetylhexosaminyltransferase (EXTL2), a member of the exostosin gene family involved in heparan sulfate biosynthesis. J. Biol. Chem., 278, 14420–14428.
Peng, L., Kawagoe, Y., Hogan, P., and Delmer, D. (2002) Sitosterol-beta-glucoside as primer for cellulose synthesis in plants. Science, 295, 59–60.
Sakaki, T., Zähringer, U., Warnecke, D.C., Fahl, A., Knogge, W., and Heinz, E. (2001) Sterol glucosides and cerebrosides accumulate in Pichia pastoris, Rynchosporium secalis and other fungi under normal conditions or under heat shock and ethanol stress. Yeast, 18, 679–695.[CrossRef][Web of Science][Medline]
Sambrook, J., Fritsch, E.F., and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, New York.
Sawardeker, J.S., Sloneker, J.H., and Jeanes, A. (1965) A. quantitative determination of monosaccharides as their alditol acetates by gas liquid chromatography. Anal Chem., 37, 1602–1605.
Saxena, I.M., Brown, R.M. Jr., Fevre, M., Geremia, R.A., and Henrissat, B. (1995) Multidomain architecture of beta-glycosyl transferases: implications for mechanism of action. J. Bacteriol., 177, 1419–1424.
Schulte, S. and Stoffel, W. (1993) Ceramide UDPgalactosyltransferase from myelinating rat brain: purification, cloning, and expression. Proc. Natl. Acad. Sci. U. S. A., 90, 10265–10269.
Sperling, P., Warnecke, D., and Heinz, E. (2004) Plant sphingolipids. In Daum, G. (ed.), Lipid Metabolism and Membrane Biogenesis. Springer, Berlin, pp. 337–381.
Sprong, H., Kruithof, B., Leijendekker, R., Slot, J.W., van Meer, G., and van der Sluijs, P. (1998) UDP-galactose: ceramide galactosyltransferase is a class I integral membrane protein of the endoplasmic reticulum. J. Biol. Chem., 273, 25880–25888.
Tadano-Aritomi, K. and Ishizuka, I. (2003) Structure and function of sulfoglycolipids in the kidney and testis. Trends Glycosci. Glycotechnol., 15, 15–27.
Tang, Y. and Hollingsworth, R.I. (1998) Regulation of lipid synthesis in Bradyrhizobium japonicum: low oxygen concentrations trigger phosphatidylinositol biosynthesis. Appl. Environ. Microbiol., 64, 1963–1966.
Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F., and Higgins, D.G. (1997) The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res., 25, 4876–4882.
Toledo, M.S., Levery, S.B., Straus, A.H., Suzuki, E., Momany, M., Glushka, J., Moulton, J. M., and Takahashi, H. K. (1999) Characterization of sphingolipids from mycopathogens: factors correlating with expression of, 2-hydroxy fatty acyl (E) -3-unsaturation in cerebrosides of Paracoccidioides brasiliensis and Aspergillus fumigatus. Biochemistry, 38, 7294–7306.[CrossRef][Medline]
Warnecke, D. and Heinz, E. (2003) Recently discovered functions of glucosylceramides in plants and fungi. Cell Mol Life Sci., 60, 919–941.[Web of Science][Medline]
Wilkinson, S.G. (1988) Gram-negative bacteria. In Wilkinson, S.G. (ed.), Microbial Lipids. Academic Press, London, Vol. 1, pp. 299–488.
Wu, K., Marks, D.L., Watanabe, R., Paul, P., Rajan, N., and Pagano, R.E. (1999) Histidine-193 of rat glucosylceramide synthase resides in a UDP-glucose- and inhibitor (D-threo-1-phenyl-2-decanoylamino-3-morpholinopropan-1-ol)-binding region: a biochemical and mutational study. Biochem. J., 341, 395–400.[Medline]
Xu, C., Fan, J., Riekhof, W., Froehlich, J.E., and Benning, C. (2003) A permease-like protein involved in ER to thylakoid lipid transfer in Arabidopsis. EMBO. J., 22, 2370–2379.[CrossRef][Web of Science][Medline]
Yanish-Perron, C., Vieira, J., and Porath, J. (1989) Improved M13 cloning vectors and host strains: nucleotide sequences of M13mp18 and pUC19 vectors. Gene, 33, 103–109.
Young, J.M., Kuykendall, L.D., Martinez-Romero, E., Kerr, A., and Sawada, H. (2001) A revision of Rhizobium Frank, 1889, with an emended description of the genus, and the inclusion of all species of Agrobacterium Conn 1942 and Allorhizobium undicola de Lajudie et al. 1998 as new combinations: Rhizobium radiobacter, R. Rhizogenes, R. Rubi, R. Undicola R. Vitis. Int. J. Syst. Evol. Microbiol., 51, 89–103.
Zähringer, U., Wagner, F., Rietschel, E.T., Ben-Menachem., G., Deutsch, J., and Rottem, S. (1997) Primary structure of a new phosphocholine-containing glycoglycerolipid of Mycoplasma fermentans. J. Biol. Chem., 272, 26262–26270.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  R. Rhome, T. McQuiston, T. Kechichian, A. Bielawska, M. Hennig, M. Drago, G. Morace, C. Luberto, and M. Del Poeta Biosynthesis and Immunogenicity of Glucosylceramide in Cryptococcus neoformans and Other Human Pathogens Eukaryot. Cell, October 1, 2007; 6(10): 1715 - 1726. [Full Text] [PDF]
|
|
|
|
 |
|
 A.-H. Lebrun, C. Wunder, J. Hildebrand, Y. Churin, U. Zahringer, B. Lindner, T. F. Meyer, E. Heinz, and D. Warnecke Cloning of a Cholesterol-{alpha}-glucosyltransferase from Helicobacter pylori J. Biol. Chem., September 22, 2006; 281(38): 27765 - 27772. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Holzl, S. Witt, A. A. Kelly, U. Zahringer, D. Warnecke, P. Dormann, and E. Heinz Functional differences between galactolipids and glucolipids revealed in photosynthesis of higher plants PNAS, May 9, 2006; 103(19): 7512 - 7517. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Holzl, U. Zahringer, D. Warnecke, and E. Heinz Glycoengineering of Cyanobacterial Thylakoid Membranes for Future Studies on the Role of Glycolipids in Photosynthesis Plant Cell Physiol., November 1, 2005; 46(11): 1766 - 1778. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||