|Home | About | Journals | Submit | Contact Us | Français|
Giardia lamblia, the protist that causes diarrhea, makes an Asn-linked-glycan (N-glycan) precursor that contains just two sugars (GlcNAc2) attached by a pyrophosphate linkage to a polyprenol lipid. Because the candidate cis-prenyltransferase of Giardia appears to be more similar to bacterial enzymes than to those of most eukaryotes and because Giardia is missing a candidate dolichol kinase (ortholog to Saccharomyces cerevisiae SEC59 gene product), we wondered how Giardia synthesizes dolichol phosphate (Dol-P), which is used to make N-glycans and glycosylphosphatidylinositol (GPI) anchors. Here we show that cultured Giardia makes an unsaturated polyprenyl pyrophosphate (dehydrodolichol), which contains 11 and 12 isoprene units and is reduced to dolichol. The Giardia cis-prenyltransferase that we have named Gl-UPPS because the enzyme primarily synthesizes undecaprenol pyrophosphate is phylogenetically related to those of bacteria and Trypanosoma rather than to those of other protists, metazoans and fungi. In transformed Saccharomyces, the Giardia cis-prenyltransferase also makes a polyprenol containing 11 and 12 isoprene units and supports normal growth, N-glycosylation and GPI anchor synthesis of a rer2Δ, srt1Δ double-deletion mutant. Finally, despite the absence of an ortholog to SEC59, Giardia has cytidine triphosphate-dependent dolichol kinase activity. These results suggest that the synthetic pathway for Dol-P is conserved in Giardia, even if some of the important enzymes are different from those of higher eukaryotes or remain unidentified.
Giardia lamblia, which is spread by the fecal–oral route, is an important parasitic cause of diarrhea in developing and developed countries (Adam 2001; Savioli et al. 2006). Giardia is remarkable for the presence of two similar nuclei and a genome that contains a large number of genes obtained from bacteria by lateral gene transfer (LGT) (Andersson et al. 2003; Franzén et al. 2009; Morrison et al. 2007).
In this report, we describe Giardia enzymes that synthesize dolichol phosphate (Dol-P), which is used to make an unusually short N-glycan precursor (dolichol pyrophosphate-GlcNAc2) and to make dolichol phosphate mannose (Dol-P-Man) (Helenius and Aebi 2004; Samuelson et al. 2005). While many eukaryotes use Dol-P-Man to make N-glycan precursors, O-linked glycans and glycosylphosphatidylinositol (GPI) anchors, Giardia only uses Dol-P-Man to make GPI anchors (Das et al. 1991; Orlean 1990; Samuelson et al. 2005).
Eukaryotic cis-prenyltransferases, which are encoded by RER2 and SRT1 genes of Saccharomyces cerevisiae, use farnesyl pyrophosphate (FPP) and numerous isopentenyl pyrophosphates (IPP) to make dehydrodolichyl pyrophosphate (Dedol-PP) (Figure 1) (Grabińska and Palamarczyk 2002; Sato et al. 1999, 2001). Dominant Dedol-PPs contain 11 isoprene units (e.g. Plasmodium, Leishmania and Trypanosoma), 16 isoprene units (e.g. S. cerevisiae Rer2p product and Trichomonas) or 19 isoprene units (e.g. S. cerevisiae Srt1p product and human) (Arruda et al. 2005; D'Alexandri et al. 2006; Grabińska et al. 2008; Löw et al. 1991; Swiezewska and Danikiewicz 2005). Dedol-PPs are subsequently dephosphorylated and then reduced to dolichol by a saturase, which has not yet been molecularly characterized (Fujii et al. 1982; Sagami et al. 1993). A cytidine triphosphate (CTP)-dependent dolichol kinase, encoded by the SEC59 gene of S. cerevisiae, converts dolichol to Dol-P (Heller et al. 1992). In bacteria, undecaprenol kinase activity is adenosine triphosphate (ATP)-dependent (Kalin and Allen 1979; Lis and Kuramitsu 2003).
Eubacteria and archaea have a cis-prenyltransferase, which makes Dedol-PP that contains 11 isoprene units (Kato et al. 1999). Like eukaryotes, archaea have a saturase that converts dehydrodolichol (Dedol) to dolichol (Burda and Aebi 1999). In contrast, eubacterial Dedol, which is used to make precursors for peptidoglycans and lipopolysaccharides, remains unsaturated (Touzé et al. 2008).
Because the candidate cis-prenyltransferase of Giardia appears to be more similar to bacterial enzymes than to most eukaryotic enzymes (see Results) and because Giardia is missing a candidate dolichol kinase (Sec59p homolog), we wondered whether Giardia synthesizes Dol-P in a manner similar to the rest of eukaryotes (Figure 1). For example, does the Giardia cis-prenyltransferase make Dedols with 11 isoprene units like those of bacteria? Does Giardia have an alternative saturase (marked in green in Figure 1), which converts Dedol-P to dolichol-P? Or does Giardia, like eubacteria, use Dedol-P for synthesis of polyprenol-linked glycans (e.g. Dedol-P-Man and Dedol-PP-GlcNAc) (alternative DPM1 and ALG7 gene products marked in blue in Figure 1)? Do Giardia membranes have CTP-dependent dolichol kinase activity, even though the parasite appears to lack a SEC59 ortholog (orange in Figure 1)?
The starting point for these studies was a characterization of the cis-prenyltransferase activity of membranes isolated from cultured Giardia. Giardia membranes make an unsaturated polyprenyl (Dedol-11), which contains 11 isoprene units when incubated with radiolabeled IPP and exogenous FPP (Figure 2). Dedol-PP phosphatase and Dedol-P phosphatase activities are inferred, because Dedol is produced by the Giardia membranes. These results are consistent with the presence of candidate cis-prenyltransferase in Giardia (see next section) (Marchler-Bauer et al. 2005).
Mass spectroscopy of polyprenols of cultured Giardia also demonstrated dolichols with 11 and 12 isoprene units (Dol-11 and Dol-12) (Figure 3A). Using the same methods, dolichols with 14 to 17 isoprene units (Dol-14 to Dol-17) were identified in Saccharomyces. The presence of Dedol-11 and Dol-11 within Giardia argues against alternative pathways for synthesis of Dol-P, which are marked in green and blue in Figure 1. Instead it appears that Giardia is synthesizing Dol-P in the same way as other eukaryotes, even though Giardia is missing a dolichol kinase candidate (SEC59 ortholog marked in orange in Figure 1) and the Giardia cis-prenyltransferase resembles those of bacteria (next section).
The predicted Giardia cis-prenyltransferase (encoded by GiardiaDB 50803_15256) is 265-amino acids long and shows a 42% identity and a 58% similarity to the Escherichia coli undecaprenyl pyrophosphate synthase (Ec-UPPS), which has been crystallized (see Supplemental Figure 1) (Guo et al. 2005). The Giardia cis-prenyltransferase, which we will refer to as Gl-UPPS because the protist synthesizes Dedol-PP with 11 isoprene units (Figure 2), contains five conserved domains that have been identified in other cis-prenyltransferases.
Phylogenetic analysis of representative eukaryotic and prokaryotic cis-prenyltransferases revealed three important findings (Figure 4). First, the vast majority of eukaryotic enzymes are present in a group (Clade 1), which is distinct from the group (Clade 2) that includes eubacteria and archaea. Second, Giardia, Trypanosoma, Leishmania and the chloroplast enzyme are also present with bacteria in Clade 2. The presence of a small number of eukaryotes in the bacterial clade is suggestive of LGT, which is a major force in the evolution of Giardia (Andersson et al. 2003; Franzén et al. 2009; Morrison et al. 2007). However, the same results could also be explained by secondary loss of the Clade 2 cis-prenyltransferase from the majority of eukaryotes. Previously, the diversity in the length of sugars present in N-glycan precursors has been shown to be secondary to loss of ALG enzymes from a common ancestor, which had a complete set (Samuelson et al. 2005).
Third, Entamoeba is the only eukaryote examined, which is missing a cis-prenyltransferase ortholog. Because the whole genomes of three different Entamoeba species have been separately sequenced (E. histolytica, E. dispar and E. invadens), the absence of a cis-prenyltransferase ortholog cannot be an artifact of the library construction (Loftus et al. 2005). This result suggests that either Entamoeba is synthesizing polyprenyls using an as yet to be identified enzyme, or Entamoeba is scavenging polyprenyls from the host.
The Giardia cis-prenyltransferase (Gl-UPPS) complements a Saccharomyces rer2Δ, srt1Δ double-deletion mutant (Figure 5 and Table I). The necessity of the Gl-UPPS is shown by the absence of growth of rer2Δ, srt1Δ double-deletion mutant when the plasmid containing the Gl-UPPS gene is lost from yeast after treatment with 1% (w/v) 5-fluoroorotic acid (FOA). Gl-UPPS in the Saccharomyces rer2Δ, srt1Δ double-deletion mutant makes polyprenols and dolichols, which contain 11 and 12 isoprene units, as described for Giardia (Figures 2 and 3, respectively). We could detect no defect in N-glycosylation (using carboxypeptidase Y (CPY) as the reporter) (Supplemental Figure 2) or in GPI anchor synthesis (using glucanosyltransferase encoded by GAS1 (Gas1p) as the reporter) (Supplemental Figure 3). While there is induction of SRT1 mRNA in the Saccharomyces rer2Δ mutant, there is no induction of SRT1 mRNA when the Saccharomyces rer2Δ mutant is complemented by the Gl-UPPS gene (Supplemental Figure 4). There is, however, a mild increase in the synthesis of chitin (consistent with mild cell wall stress) in Saccharomyces rer2Δ, srt1Δ double-deletion mutant complemented with Gl-UPPS (Supplemental Figure 5).
When the Gl-UPPS with a histidine tag was expressed in E. coli, the recombinant protein, which was purified using a nickel column, made prenol-11-PP when incubated with radiolabeled IPP and exogenous FPP (Figure 2). This result suggests that the activity of the Gl-UPPS is not dependent upon accessory proteins.
Using the Saccharomyces Sec59p as a probe, we were able to identify candidate dolichol kinases from all eukaryotes examined with the exception of Giardia (Supplemental Figure 6). As whole genome sequences were examined from two different Giardia isolates (WB and GS), it is unlikely that the absence of a Giardia dolichol kinase ortholog is due to an artifact in library construction or sequencing (Franzén et al. 2009; Morrison et al. 2007). In addition, we were unable to find in Giardia orthologs to S. cerevisiae diacylglycerol kinase (Dgk1p) (Han et al. 2008), Arabidopsis thaliana phytol kinase (Valentin et al. 2006), Bacillus subtilis undecaprenol kinase or E. coli diacylglycerol kinase (Lis and Kuramitsu 2003). Indeed recent work clearly demonstrates that the purified E. coli bacA protein (homolog of B. subtilis undecaprenol kinase) exhibits undecaprenyl pyrophosphate phosphatase activity but not undecaprenol kinase activity (El Ghachi et al. 2004).
Nevertheless, Giardia membranes have CTP-dependent dolichol kinase activity, which is increased when exogenous dolichol is added (Figure 6). Giardia membranes have no ATP-dependent dolichol kinase activity, as has been described for the bacterial undecaprenol kinase (Kalin and Allen 1979; Lis and Kuramitsu 2003). Similarly, Giardia membranes have no uridine triphosphate (UTP)- or guanosine triphosphate (GTP)-dependent dolichol kinase activity in vitro. We conclude that Giardia either has a deeply divergent dolichol kinase, which was not detected using the Sec59p probe or Giardia has a unique dolichol kinase that does not share common ancestry with Sec59p. We cannot rule out the possibility that the Giardia dolichol kinase came from bacteria by LGT, but we have no evidence for this.
Major conclusions include the following:
Trophozoites of the first genome project WB strain of Giardia were grown axenically in TYI-S media supplemented with 10% serum and 1 mg/mL bile (Morrison et al. 2007). Giardia cells were chilled on ice for 20 min and then concentrated by centrifugation.
Saccharomyces strains used in this study, which include single deletion strains made on a BY4743 background that were obtained from Euroscarf, are listed in Table I (Brachmann et al. 1998). Yeasts were cultured in 2% (wt/vol) Bacto peptone and 1% (wt/vol) yeast extract supplemented with 2% glucose (wt/vol) (YPD, yeast peptone dextrose medium). Synthetic minimal media were made of 0.67% (wt/vol) yeast nitrogen base and 2% (wt/vol) glucose, supplemented with auxotrophic requirements. For solid media, agar (Difco, Voigt Global Distribution Inc, Lawrence, KS) was added at a 2% (wt/vol) final concentration. Sporulation of the diploid cells and tetrad dissection were performed by standard yeast genetic methods. Yeast cells were grown at 30°C and harvested at logarithmic growth phase (1 to 2 OD units/mL).
The dolichol fraction was isolated from membrane of Saccharomyces (10 mg of protein) and Giardia (40 mg of protein), as described (Grabińska et al. 2005) and subjected to analysis by liquid chromatography and mass spectrometry (LC-MS). To increase the amount of dolichol in the Giardia sample before separation of the lipids on the silica gel column, we treated prenyl phosphates with potato acid phosphatase, as described (Fujii et al. 1982).
LC/MS of lipids was performed using a Shimadzu LC system (comprising a solvent degasser, two LC-10A pumps and a SCL-10A system controller) coupled to a QSTAR XL quadrupole time-of-flight tandem mass spectrometer (as above). LC was operated at a flow rate of 200 μL/min with a linear gradient as follows: 100% of mobile phase A was held isocratically for 2 min and then linearly increased to 100% mobile phase B over 14 min and held at 100% B for 4 min. Mobile phase A consisted of methanol/acetonitrile/aqueous 1 mM ammonium acetate (60/20/20, v/v/v). Mobile phase B consisted of 100% ethanol containing 1 mM ammonium acetate. A Zorbax SB-C8 reversed-phase column (5 μm, 2.1 × 50 mm) was obtained from Agilent (Palo Alto, CA). The postcolumn splitter diverted +AH4-10% of the LC flow to the electrospray ionization source of the mass spectrometer.
The LC-MS results were compared with an in vitro cis-prenyltransferase assay performed as described (Szkopinska et al. 1997). Briefly, Saccharomyces or Giardia membranes were incubated with [1-14C] IPP (60 mCi/mmol) and exogenous FPP. Gl-UPPS activity was stimulated by the presence of 0.1% Triton X-100. The length of the Dedol product was determined by reverse-phase thin layer chromatography (TLC) using standards produced in vitro by BY4741 strain of Saccharomyces and commercially available undecaprenols (American Radiolabeled Chemicals, Inc., St Louis, MO). Giardia membranes were prepared according to procedures described for preparation of Trichomonas membranes (Grabińska et al. 2008).
To make recombinant Gl-UPPS and recombinant E. coli bacA undecaprenyl pyrophosphate phosphatase in E. coli, coding sequences of each were amplified with polymerase chain reaction (PCR) and cloned into the pET30a vector (EMD Biosciences-Novagen, Madison, WI) in such a way that each protein was tagged at the N-terminus with a polyhistidine-S-tag. pET30a vectors containing the Gl-UPPS or bacA genes were each transformed into E. coli Rosetta 2 cells (Novagen). E. coli in the logarithmic growth phase were induced to express heterologous protein by incubation with 1 mM isopropyl-β-d-thiogalactopyranoside for 4 h at 30°C. The harvested cells were lysed by sonication, and His-tagged proteins were purified on a nickel column according to the Invitrogen protocol (Gl-UPPS) or by published methods (El Ghachi et al. 2004). Protein purity was judged by sodium dodecyl sulfate–polyacrylamide gel electrophoresis and western blotting using a monoclonal mouse antibody against the S tag.
The cis-prenyltransferase assay was done using 10 μg of Gl-UPPS, and Dedol-PP product was dephosphorylated with undecaprenyl pyrophosphate phosphatase, as described (El Ghachi et al. 2004). Products were separated by TLC on precoated plates of silica gel 60 (Merck) using diisobutyl ketone/acetic acid/water (8:5:1, v/v/v) as a mobile phase.
The predicted proteins of Giardia of the WB strain (first genome project) and GS strain (second genome project), which have been deposited in the non-redundant (NR) data at GenBank or GiardiaDB, were searched with Psi-Blast using cis-prenyltransferases (Rer2p and Srt1p) and dolichol kinase (Sec59p) sequences from Saccharomyces (Altschul et al. 1997; Aurrecoechea et al. 2008; Franzén et al. 2009; Heller et al. 1992; Morrison et al. 2007; Sato et al. 1999, 2001). Giardia proteins were also searched with B. subtilis undecaprenol kinase and E. coli diacylglycerol kinase (Lis and Kuramitsu 2003). Similar methods were used to search the predicted proteins of representative protists, metazoans, fungi, plants, eubacteria and archaea in the NR database at the National Center for Biotechnology Information (NCBI).
The single predicted cis-prenyltransferase of the WB strain of Giardia (GiardiaDB 50803_15256 or GenBank EDO82194) was examined for conserved domains using the CD search at the NCBI (Marchler-Bauer et al. 2005). The set of eukaryotic and prokaryotic cis-prenyltransferases was aligned using multiple sequence comparison by log-expectation (Edgar 2004). The alignment was manually refined, and gaps were removed using BioEdit. The finished alignment was used to construct the phylogenetic tree using TREE-PUZZLE, a program to reconstruct phylogenetic trees from molecular sequence data by the maximum likelihood method (Schmidt et al. 2002). Similar methods were used to draw the Sec59 tree.
A Saccharomyces rer2Δ/RER2, srt1Δ/SRT1 double-deletion mutant (KG219) was made using a single Saccharomyces rer2Δ/RER2 deletion strain (Y23137) as a starting point. Briefly, deletion of SRT1 gene was accomplished using the plasmid pUG27 that carries the loxP-his5+-loxP gene disruption cassette (Gueldener et al. 2002). PCR primers used to target the srt1 gene were TTATAAAGAACAGGCTGCCTTTCAAACATAGGACGTTTCTGTTGACCATACAGCTGAAGCTTCGTCTTCGTACGC (sense) and TTCAGAATGTTCTTTGCCCTCTCTTGGGCTTTCTAGTTTTGCACTTTTACGCATAGGCCACTAGTGGATCTG (antisense). The pUG27-srt1 knock-out construct was transformed into Y23137 yeast cells. Transformants able to grow on medium lacking histidine and containing G418 were isolated, and the correct insertion of the deletion cassette was verified by PCR. The double heterozygous mutant rer2::kanMX4/RER2, srt1:: his3MX6loxP/SRT1 was called KG219.
The coding sequence of putative Giardia cis-prenyltransferase, which was fused to the EKKL ER-retention signal, was amplified from WB strain genomic DNA using two custom primers. The sense primer, which contained a HindIII restriction enzyme site in bold, was AAAAAGCTTATGATCCCCATGCATGTGGC. The antisense sequence, which contained a SalI restriction site in bold and encoded EKKLN in italic, was TTTGTCGACTCAATTCAACTTTTTTTCGTCATGGTTCGATAGGG. The PCR product was cloned into the pGEM-T Easy vector and sequenced (Promega, Madison, WI).
To express Giardia cis-prenyltransferase (Gl-UPPS) in the Saccharomyces cells, a NotI-surrendered insert was subcloned into the pNEV-N plasmid under the control of the PMA1 promoter and terminator (Sauer and Stolz 1994). The pNEV–Gl-UPPS plasmid was transformed into the KG219 yeast strain. Yeasts were sporulated, and colonies were selected that were resistant to G418 and able to grow on the medium lacking histidine. These yeasts, which express the Giardia cis-prenyltransferase in a Saccharomyces rer2Δ, srt1Δ double-deletion mutant, were labeled KG119. A single deletion mutant rer2Δ that expressed GL-UPPS was called KG120.
KG119 and its wild-type counterpart BY4741 were characterized in six ways. 1) Polyprenols were extracted and characterized by mass spectroscopy, as described above for Giardia. 2) Membranes were isolated and incubated with radiolabeled IPP and exogenous FPP, and radiolabeled Dedols were demonstrated by reverse-phase TLC, as described above. 3) The N-glycosylation status of carboxypeptidase (CPY) was determined by western blotting of Saccharomyces glycoproteins with antibodies to CPY (Molecular Probes, Eugene, OR) before or after peptide:N-glycanaseF treatment. A negative control was a Saccharomyces alg5Δ mutant (Heesen et al. 1994). 4) Maturation of the GPI anchors present on the Saccharomyces Gas1p (Gatti et al. 1994) was determined by western blotting of Saccharomyces glycoproteins with antibodies to Gas1p (kind gift of Laura Popolo). A negative control was a Saccharomyces gpi1 thermosensitive mutant (Leidich et al. 1994). 5) Expression of SRT1 mRNAs was measured in wild type versus Saccharomyces rer2Δ with or without the exogenous Gl-UPPS. 6) Cell wall stress was indirectly determined by measuring chitin levels (Popolo et al. 1997). Chitin content was measured by an assay adapted for microtiter plates, as described (Grabińska et al. 2007).
CTP, GTP and UTP were synthesized enzymatically from cytidine diphosphate, guanosine diphosphate or uridine diphosphate and [γ-32P]ATP with nucleoside-5′-diphosphate kinase, as described (Han et al. 2008). The dolichol kinase assay was performed, as described (Heller et al. 1992). Membrane fractions (200 μg) were incubated in a total volume of 100 μL containing 0.05 M Tris–HCl (pH 7.5), 10 mM UTP, 100 mM CaCl2, 0.02–0.1 μCi [γ-32P]CTP in 0.1% Triton X-100 and 2 μg of dolichol mixture for 30 min at room temperature. Alternatively, [γ-32P]ATP or [γ-32P]GTP was used instead of [γ-32P]CTP. When [γ-32P]UTP substituted for [γ-32P]CTP, then 10 mM ATP was used instead of 10 mM UTP. Reactions were terminated by the addition of 750 μL of 1 M KOH in methanol, and alkali-labile lipids were hydrolyzed by incubation at 37°C for 30 min. This step is required to hydrolyze phosphatidic acid. The lipids were extracted by the Folch method, and the 32P incorporation into Dol-P was determined by scintillation counting.
Supplementary data for this article is available online at http://glycob.oxfordjournals.org/.
National Institutes of Health grants AI048082 (to J.S.) and GM31318 (to P.W.R.). Z.G. and the mass spectrometry facility in the Department of Biochemistry, Duke University Medical Center were supported by a LIPID MAPS glue grant (GM-069338) from the National Institutes of Health.