Glycobiology Advance Access originally published online on February 5, 2008
Glycobiology 2008 18(4):303-314; doi:10.1093/glycob/cwn008
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Engineering of a mammalian O-glycosylation pathway in the yeast Saccharomyces cerevisiae: production of O-fucosylated epidermal growth factor domains
2 Research Institute for Cell Engineering, National Institute of Advanced Industrial Science and Technology (AIST), 1-1-1, Higashi, Tsukuba, Ibaraki 305-8566
3 Department of Applied Molecular Biosciences, Nagoya University Graduate School of Bioagricultural Sciences, Furo-cho, Chikusa-ku, Nagoya 464-8601, Japan
1 To whom correspondence should be addressed: Tel: +81-29-861-6160; Fax: +81-29-861-6161; e-mail: jigami.yoshi{at}aist.go.jp
Received on June 6, 2007; revised on January 29, 2008; accepted on January 31, 2008
 |
Abstract |
|---|
 Top Abstract IntroductionResultsDiscussionMaterials and methodsFundingConflict of interest statementReferences |
|---|
Development of a heterologous system for the production of homogeneous sugar structures has the potential to elucidate structure–function relationships of glycoproteins. In the current study, we used an artificial O-glycosylation pathway to produce an O-fucosylated epidermal growth factor (EGF) domain in Saccharomyces cerevisiae. The in vivo O-fucosylation system was constructed via expression of genes that encode protein O-fucosyltransferase 1 and the EGF domain, along with genes whose protein products convert cytoplasmic GDP-mannose to GDP-fucose. This system allowed identification of an endogenous ability of S. cerevisiae to transport GDP-fucose. Moreover, expression of EGF domain mutants in this system revealed the different contribution of three disulfide bonds to in vivo O-fucosylation. In addition, lectin blotting revealed differences in the ability of fucose-specific lectin to bind the O-fucosylated structure of EGF domains from human factors VII and IX. Further introduction of the human fringe gene into yeast equipped with the in vivo O-fucosylation system facilitated the addition of N-acetylglucosamine to the EGF domain from factor IX but not from factor VII. The results suggest that engineering of an O-fucosylation system in yeast provides a powerful tool for producing proteins with homogenous carbohydrate chains. Such proteins can be used for the analysis of substrate specificity and the production of antibodies that recognize O-glycosylated EGF domains.
Key words: EGF domain / Fringe / O-fucose / protein O-fucosyltransferase 1 / yeast
|
Introduction |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsFundingConflict of interest statementReferences |
|---|
Glycosylation is important for many of the biological functions of glycoproteins. Because the biosynthesis and regulation of glycosylation are complex, the production of therapeutic glycoproteins is usually done using cells from higher eukaryotes, such as mammalian cell lines. However, the use of mammalian cell lines for glycoprotein production has several drawbacks, including low protein production, long culture time, heterogeneous products, and viral contamination issues.
The yeast Saccharomyces cerevisiae is a well-established organism that can be grown to a high cell density within a few days in a chemically defined medium and can easily be genetically modified. Moreover, yeast cells possess the relevant organelles (i.e., endoplasmic reticulum [ER] and Golgi) necessary for the biosynthesis of glycoproteins. However, the specific sugar components and structures used in the production of yeast glycoproteins differ from those of mammalian cells (Nakajima and Ballou 1974a
, 1974b). These differences sometimes cause glycoproteins endogenously or heterologously produced in yeast to be immunogenic in humans. Nonetheless, S. cerevisiae has the advantage that it can be used for precise studies of the structure–function relationship of glycoproteins, as S. cerevisiae is a robust system for the production of homogeneous sugar structures via molecular genetic manipulation. Furthermore, a heterologous system in yeast could be used to identify essential components of the donor cell system by excluding the effects of cell-specific factors not present in host cells.
Several attempts have been made to produce mammalian N-linked glycoproteins in yeast by expressing the genes that carry out mammalian-specific glycosylation (Chiba et al. 1998; Choi et al. 2003; Li et al. 2006). To date, however, no attempts have been made to reconstitute O-linked mammalian glycosylation pathways, which normally occur in the relevant organelles of those cells, in a eukaryotic microorganism. We have been interested in reconstituting the O-glycosylation pathway in yeast, and in the current studies, we focused our attention on the production of heterologous glycoproteins containing fucose. Because eukaryotic microorganisms, including yeast and fungi, generally do not contain a biosynthetic pathway for GDP-fucose production, such a pathway must be provided in addition to the genes that encode proteins required for O-fucosylation itself. We previously succeeded in efficiently generating cytoplasmic GDP-fucose in S. cerevisiae from endogenous GDP-mannose via coexpression of Arabidopsis thaliana MUR1, which encodes GDP-mannose-4,6-dehydratase, and AtFX/GER1, which encodes GDP-4-keto-6-deoxy-mannose-3,5-epimerase/4-reductase (Nakayama et al. 2003).
Glc-Fuc-O-Thr, a compound in human urine, was the first O-fucosylated structure to be identified (Hallgren et al. 1975). Subsequently, O-linked fucose was found to be attached to the epidermal growth factor (EGF) domain of recombinant urokinase (Kentzer et al. 1990). Recently, the O-fucosylated structure of the EGF domain has been the subject of a considerable study, as it is essential for the function of a urokinase-type plasminogen activator (Rabbani et al. 1992) and for Notch signaling (Okajima and Irvine 2002; Shi and Stanley 2003). The EGF domain consists of approximately 40 amino acids and contains three disulfide bonds formed by six conserved cysteine residues. The O-fucose is attached by O-fucosyltransferase 1 (O-FucT-1) to the serine or threonine residue adjacent to the third conserved cysteine within the consensus sequence C2X4-5(S/T)C3 (Panin et al. 2002; Shao et al. 2003). Wang et al. reported that a correctly folded EGF domain containing the three conserved disulfide bonds is important for recognition by O-FucT-1 (Wang et al. 1996; Wang and Spellman 1998). The native O-fucosylated structure of the EGF domain can be classified into two types (Bjoern et al. 1991; Nishimura et al. 1992; Harris and Spellman 1993): (i) a fucosylated monosaccharide structure and (ii) a further elongated tetrasaccharide structure, NeuAc
2-6Galβ1-4GlcNAcβ1-3Fuc1, in which the O-linked fucose is a substrate for β1,3-N-acetylglucosaminyltransferases encoded by fringe genes (Bruckner et al. 2000; Moloney et al. 2000). The results of previous reports suggest that the differences in O-linked sugar chain structures are caused by specificity of the Fringe protein (Shao et al. 2003; Rampal et al. 2005).
Here, we report the first example of reconstitution in yeast of a mammalian O-fucosylation system (Figure 1A). Efficient O-fucosylation of a secreted EGF domain was confirmed by HPLC and matrix-assisted laser desorption ionization-time of flight (MALDI-ToF) MS, and by fucose-specific lectin blotting. To investigate the specificity of O-FucT-1, we expressed mutant forms of the EGF domain predicted to affect the formation of disulfide bridges in the engineered yeast cells. Furthermore, we demonstrate that the fucosylated EGF domain derived from human factor IX but not from factor VII acts as an acceptor for the in vivo addition of GlcNAc by human Manic Fringe expressed in yeast.
|
|
Results |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsFundingConflict of interest statementReferences |
|---|
Construction of an in vivo O-fucosylation system in yeast
With the aim of constructing a heterologous system for the production of O-fucosylated mammalian proteins, we first attempted to clone the mammalian genes necessary for in vivo O-fucosylation and express them in a functional form in yeast. To do this, the VSVG-tagged human GDP-fucose transporter gene (hGFR) (Lubke et al. 2001
; Luhn et al. 2001) and Flag-tagged human O-FucT-1 coding sequences (POFUTI) were expressed in yeast cells that produce GDP-fucose in the cytosol from intrinsic GDP-mannose via the activity of heterologously expressed MUR1 and AtFX. The hGFR and POFUT1 genes were integrated at the URA3 or LEU2 locus of the yeast chromosome and expression was confirmed by immunoblotting (Figure 1B). We used the EGF domain-1 derived from human factor VII as an acceptor for in vivo O-fucosylation, as it was the first example of an acceptor for O-FucT-1 (Wang et al. 2001; Wang and Spellman 1998). To produce the EGF domain, an expression cassette containing (1) a codon-optimized coding sequence that encodes a single EGF domain, (2) a hexahistidine tag, and (3) the yeast -factor pre-pro sequence was introduced into the yeast strain W303-1B and integration at the TRP1 locus was confirmed (Figure 2A). The secreted EGF domains were purified using Ni-NTA agarose from cultures of cells containing all of the necessary genes for in vivo O-fucosylation. Expression was confirmed by immunoblotting and furthermore, SDS–PAGE followed by staining with Coomassie brilliant blue showed that the protein was efficiently produced and secreted by the engineered yeast (Figure 2B). Amino acid sequencing revealed that the N-terminal amino acid sequence of the EGF domain was Ala-Gly-His-His-His-His-His-His-Val-Ser-Asp, confirming that as expected, the protein was cleaved by Kex2p after Lys-Arg.
|
To confirm that fucose is added to the EGF domain in this system, we purified the protein by HPLC (Figure 2C). A new peak (b), which is likely to have a higher polarity than the original peak (a), was observed in the sample derived from engineered yeast cells (Figure 2C, panels 1 and 2). We collected the peaks (a) and (b) and analyzed them by MALDI-ToF MS (Figure 2D). The observed mass of peak (a) was 5703.6. This is nearly identical to the predicted mass of 5704.1, implying the formation of three disulfide bonds. The new peak (b) had a mass of 5849.8, which is 146.2 higher than that of peak (a). This difference is identical to the theoretical mass of a deoxyhexose, fucose (146.1), implying the addition of a single fucose to the recombinant EGF domain. In this system, the presence of a significant amount of fucosylated EGF domain in the secreted EGF domain strongly suggests that all of the proteins necessary for O-fucosylation (GDP-fucose transporter, O-FucT-1, and two enzymes for converting GDP-mannose to GDP-fucose) are functional. Moreover, it suggests that the EGF domain is folded in such a way that it is recognized by O-FucT-1 in yeast cells. These results demonstrate successful in vivo O-fucosylation of the EGF domain in engineered yeast cells.
Optimization of O-fucosylation efficiency
To optimize the engineered yeast system, we further modified the yeast strains. Because several nutritional marker genes were already occupied by recombinant genes in the engineered yeast cells (YCY9 cells, see Table I), further introduction and selection of heterologous genes would be a difficult challenge. In addition, YCY9 cells grow slower than the parental cells. We reasoned that this could be due to the fact that the heterologous membrane protein (hGFR) is overproduced in the yeast Golgi membrane in these cells. Recently, it was reported that nucleotide sugar transporters (NST) have broad but not very sharp preference to nucleotide sugars (Berninsone et al. 2001; Hong et al. 2000; Segawa et al. 2002). Therefore, we made the reasonable assumption that the endogenous GDP-mannose transporter might transport GDP-fucose in S. cerevisiae. To test this, we prepared cells containing MUR1, AtFX, O-FucT-1, and the EGF domain but lacking hGFR (TOY10). Unlike YCY9 cells, TOY10 cells had a growth rate similar to that of parental cells. Interestingly, the O-fucosylated form of the EGF domain was detected in TOY10 cell cultures (Figure 3A, -KPi, TOY10). O-Fucosylation in TOY10 cells indicates that S. cerevisiae has endogenous GDP-fucose transport ability, despite the apparent lack of machinery for protein fucosylation in yeast.
|
|
In this assay, we also found that yeast culture conditions significantly affect the amount of EGF domain production and the ratio of O-fucosylated to unmodified EGF domain secreted into the culture medium. HPLC analysis revealed that the production of the EGF domain increased when a potassium phosphate buffer was added to the medium (+KPi, Figure 3A). The total amount of O-fucosylated EGF domain gradually increased during cultivation, and most of the secreted domain was fucosylated after 4 days when cells were cultured in a YPADP medium (+KPi, TOY10, Figure 3A). By contrast, when cells were cultured in the YPAD medium, no significant change over time in the proportion of fucosylated domain was observed (-KPi, TOY10, Figure 3A). Moreover, under those conditions, the total amount of both products decreased, presumably due to the degradation by acidic proteinases secreted into the culture medium.
To address the possibility that efficient O-fucosylation might occur in the medium after cell lysis, TOY10 cells were grown in the YPADP medium for 1, 2 or 3 days and cells were removed by centrifugation, then the culture broth was further incubated for 24 h at 30°C (Figure 3B). The peak ratio was not changed for a 1-day culture sample with a further incubation of 1 day (Figure 3B, 1–2 days). In contrast, the amount of O-fucosylated EGF domain increased for 2- and 3-day culture samples with a further 1-day incubation (Figure 3B, 2–3 days, 3–4 days), suggesting that some of O-FucT-1 and GDP-fucose are released from the cells into the culture medium by secretion or cell lysis, resulting in O-fucosylation in the media. The amount of O-fucosylated EGF domain peak was largest when a 3-day culture medium was incubated for another 1 day (Figure 3B, 3–4 days). Further to address the degree of O-fucosylation that occurs in the media relative to the total amount of O-fucosylated product, we prepared both YCY3 cells harboring the hPOFUT1 gene and YCY8 cells harboring the EGF domain gene and the MUR1 and AtFX genes, which were grown together in the YPADP medium (coculture; Figure 3C). First, when we performed the coculture starting at a 1/1 seed ratio for YCY3/YCY8 cells, the total amount of secreted EGF domain from YCY8 cells decreased to less than 10% of that of secreted domain from TOY10 cells (data not shown), probably due to the slower growth rate of YCY8 cells than that of YCY3 cells under the coculture system. Then, we tried to coculture YCY3/YCY8 cells at a seed ratio of 1/5. The amount of secreted EGF domain from YCY8 cells in a coculture medium increased to 
40% of that from TOY10 cells. The ratio of O-fucosylated EGF to total EGF domains secreted into a coculture medium was 510% for 2 days and about 40% for 4 days (Figure 3C). The amount of O-fucosylated domain was similar when the seed ratio of TCY3/YCY8 cells was changed to 1/1 or 1/2 (data not shown), when a sufficient amount of O-FucT-1 was supplied from YCY3 cells. In contrast, the ratio of O-fucosylated EGF to total EGF domains was 30–40% for 2 days and this ratio increased to 90% for 4 days in the culture medium of TOY10 cells (Figure 3A, +KPi, 2 days, TOY10, and +KPi, 4 days, TOY10).
These results suggest that O-fucosylated domains were produced inside the cells at the early culture period of TOY10 cells and O-fucosylation in the culture medium increased at the later stage of cultivation, presumably due to the combination of secretion of O-FucT-1 and the supply of GDP-fucose by cell lysis, resulting that about 50% of O-fucosylated domain would be formed outside the cells after 4 days.
We further examined the intracellular O-fucosylation by the direct approach. The O-fucosylated product was detected on HPLC analysis of the purified EGF domain obtained from fresh extracts of TOY10 cells, but not from those of YCY4 cells (Figure 3D). The above intracellular EGF domains separated by HPLC were further analyzed on its molecular mass by MALDI-ToF MS. The molecular mass of m/e = 5850.0, which is identical with the calculated mass value for the O-fucosylated EGF domain (m/e = 5849.7), was detected in extracts of TOY10 cells, but not of YCY4 cells (data not shown). Therefore, we confirmed the significant O-fucosylation inside the cells.
Taken together, the data show that not introducing a recombinant GDP-fucose transporter and optimizing culture conditions were very much effective for the efficient production of the O-fucosylated EGF domain.
Expression of EGF domain mutants in the engineered yeast cells
Wang et al. (Wang et al. 1996; Wang and Spellman 1998) reported that disulfide pairing of the EGF domain and correct folding are important to the ability of the domain to serve as an acceptor substrate for O-FucT-1, as short synthetic peptides derived from the EGF domain or replacement of either or both glycine residues of the EGF domain consensus sequence are poor acceptors for the substrate. To determine the role of disulfide pairing in folding and acceptor recognition in our system, we constructed three EGF domain mutants based on the coding sequence of the EGF domain-containing factor VII protein. In the mutants, each one of three pairs of cysteines was replaced with a pair of alanine residues (Figure 4A).
|
HPLC analysis revealed differences in the peak patterns on the chromatograms of the unmodified and mutant-form domains (Figure 4, B–I). To further investigate this effect, we performed MS analysis on each relevant HPLC peak fraction. The observed masses revealed that the peaks with the elution times indicated in Figure 4 represent polypeptides derived from full-length domain mutants, with the exception of a peak at 14.2 in the 2-4 mutant (data not shown). Although the chromatogram of the 2-4 mutant showed multiple peaks, the overall level of the 2-4 mutant domain was comparable to the levels of the unmodified EGF domain (Figure 4, H versus F). By contrast, the 1-3 and 5-6 mutant peaks were much smaller than that of the unmodified form (Figure 4, G and I versus F). This suggests that a lack of C2–C4 disulfide pairing has a little effect on the secretion and/or stability of the EGF domain but that C1–C3 and C5–C6 disulfide pairings are critical for efficient secretion and/or stability of the EGF domain.
It is notable that on the MS analysis, it appears that the 1-3 mutant polypeptide fraction contained a small but detectable amount of fucose. The results obtained for the 24.0 peak from the 1-3 mutant, in which a difference of 146.4 is observed relative to the main peak, are consistent with the presence of a deoxyhexosyl fucose residue (Figure 5A). On the other hand, no such difference of 146.4 is detected from the 2-4 and 5-6 mutant polypeptide fractions by MS analysis. Furthermore, we tested for the attachment of a fucose residue to EGF domain mutants by Aleuria auramtia lectin (AAL) blotting (Figure 5B). In this assay, only the 1-3 mutant EGF domain yielded a signal, implying the presence of fucose on that mutant form. Consistent with MS data, this result suggests that C1–C3 disulfide pairing is not essential for the fucosylation of the EGF domain. It is likely that the 1-3 mutant folded form is similar in structure to the wild-type EGF domain, such that it can act as an acceptor for O-FucT-1, provided that the other C2–C4 and C5–C6 disulfide bonds do form.
|
Recognition of O-fucosylated EGF domains by fucose-specific lectin
We also prepared an EGF domain-encoding construct from human factor IX and expressed it in our yeast system. Factor VII contains fucose alone (Bjoern et al. 1991
), whereas factor IX involves an additional trisaccharide attachment to the O-fucosylated protein (Nishimura et al. 1992). SDS–PAGE and HPLC analysis of the EGF domain derived from factor IX showed that the domain is produced and secreted by the engineered yeast cells, similar to what was observed for the factor VII domain (data not shown). O-Fucosylated and non-fucosylated EGF domains from factor IX could not be separated by HPLC even after tests were performed under several different conditions (data not shown). However, the results of MS analysis (Figure 6A) reveal the presence of an O-fucosylated domain (IX+Fuc) in the engineered yeast cells, as the observed mass (5684.6) was 145.9 higher than the observed mass of the domain when produced in yeast lacking the O-fucosylation system (5538.7; IX only).
|
We next used lectin blotting to ask if the O-fucosylated structures attached to the EGF domains of factors VII and IX expressed in yeast are recognized by fucose-binding lectin. The assay was useful to examine the role of disulfide bonds in recognition of O-fucosylated structures by lectins, as SDS–PAGE was carried out in both reducing and nonreducing conditions prior to the binding assay. Under reducing conditions, lectin blotting revealed a clear signal with the EGF domain from factor VII but not from factor IX, and no signal was detected for either under nonreducing conditions (Figure 6B). These results suggest that AAL binds to the O-fucosylated structure of the EGF domain from factor VII but binds very weakly or not at all to the EGF domain from factor IX under denaturing, reducing conditions.
Notably, under nonreducing conditions mobility on SDS–PAGE of the EGF domain from factor VII (EGFVII) was different from that of the O-fucosylated EGF domain (EGFVII+Fuc), whereas mobility of the EGF domain from factor IX was the same, irrespective of fucosylation (EGFIX and EGFIX+Fuc; Figure 6B). Different mobility on SDS–PAGE suggests the differences in the binding of SDS between the factor VII EGF domain and the factor IX domain. Moreover, it seems reasonable to suggest that the effect is due to a difference in the three-dimensional structures of the O-fucosylated EGF domains of factors VII and IX.
Manic fringe induces elongation of O-fucose in the engineered yeast cells
We next attempted to express human Manic Fringe in the yeast system and asked if expression of this gene results in the elongation of the O-fucosylated domain. A VSVG-tagged Manic Fringe coding sequence was introduced via an integration plasmid into engineered yeast cells that produce human factor VII or IX O-fucosylated EGF domains. Integration of the Manic Fringe construct at the URA3 locus was confirmed by PCR and expression of each exogenously introduced gene was confirmed by immunoblotting (data not shown). To assay elongation of the sugar, EGF domains were purified from the culture broth of engineered yeast and the molecular mass of the proteins was analyzed by MALDI-ToF MS after fractionation via HPLC.
The MS spectra of the fractionated factor IX EGF domain showed three peaks (peaks 0, 1 and 2; Figure 7A, panel I). Peak 0 had an observed mass of 5539.5, consistent with the secreted EGF domain lacking a sugar component (theoretical mass of 5538.9). Peaks 1 and 2 have higher observed masses (5685.5 and 5888.0, respectively). The mass of peak 1 is consistent with an EGF domain with O-fucose. Two pieces of evidence suggest that the mass of peak 2 is consistent with the further addition of N-acetylhexosamine (HexNAc). First, the observed difference in mass was 202.5, consistent with the predicted mass of 203.1 for HexNAc. Second, peak 2 decreased dramatically after in vitro treatment with N-acetylhexosaminidase (Figure 7A, panel II).
|
To identify the HexNAc species, we performed a monosaccharide analysis using a p-aminobenzoic acid ethylester (ABEE)-labeled sugar derivative after the acid hydrolysis via HPLC fractionation (Figure 7B). The hydrolysis product from the EGF domain secreted by the engineered yeast cells in the absence of the heterologous Manic Fringe construct showed a fucose peak alone (Figure 7B, YCY12). However, the hydrolysis product from the cells harboring Manic Fringe showed a GlcNAc peak in addition to the fucose peak, whereas GalNAc peak, which would elute in
40 min, was not detectable (Figure 7B, YCY16). This result strongly suggests that the HexNAc species is GlcNAc. Interestingly, a GlcNAc peak was not detected in the MS spectra or via monosaccharide analysis of the EGF domain from factor VII (data not shown), suggesting that the O-fucosylated structure of factor IX but not of factor VII acts as a substrate for human Manic Fringe in yeast cells. |
Discussion |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsFundingConflict of interest statementReferences |
|---|
Glycans play essential roles in protein folding, quality control, sorting, and transport in glycoprotein-producing cells (Helenius and Aebi 2001
). Among several sugar components, fucose is particularly important to mammalian glycoproteins as it allows an increase in the number of branches in the carbohydrate structure and, thus, diversity in N-glycans. Moreover, this diversity extends to the antigen structure of mucin O-glycans in tumor cells (Kui Wong et al. 2003). Fungi and yeast, including S. cerevisiae, lack the ability to carry out protein fucosylation. Thus, it has not previously been feasible to synthesize mammalian-type glycans that contain fucose in yeast. However, a system for the production of glycans in yeast would be useful, as yeast have advantages over mammalian cell culture systems in terms of large-scale production, rapid growth, and low-cost cultivation. Here, we have reported the generation of yeast cells that express plant and human genes necessary for early steps in O-fucosylation in metazoa. To the best of our knowledge, this is the first report of the heterologous production of a mammalian O-fucosylated structure in yeast cells. In the study, we confirmed that the secreted O-fucosylated peptide was produced by the engineered yeast cells in quantities sufficient for many applications (5 mg/L culture), and without competition from the endogenous O-mannosylation pathway. The production level of the EGF domain and the ratio of O-fucosylated to unmodified significantly increased by using a culture medium in which pH was controlled by the addition of a potassium phosphate buffer. As a part of the O-fucosylated domain, GDP-fucose released into the culture medium after cell lysis is likely to be transferred to the secreted EGF domain via extracellular O-FucT-1. However, the percentage of O-fucosylated domain in the culture medium is estimated to be at most about 50% of the total amount of O-fucosylated product after 4 days culture. Most of the O-fucosylated domains were derived from inside the cell at the early stage of cultivation, as supported by both the coculture experiment, which showed a small amount of O-fucosylated EGF domain in the culture medium and the analysis of the purified EGF domain in freshly prepared cellular extracts of the engineered cells cultured in the YPADP medium. Whatsoever, our system combined with in vivo and ex vivo O-fucosylation by yeast cells can be used to produce peptides or proteins containing homogenous mammalian O-fucosylated structures for further applications; i.e., for the production of antigens useful for the generation of antibodies, or modified proteins that can be crystallized for X-ray analysis.
MS analysis of the O-fucosylated product clearly indicates the presence of three disulfide bridges, formed by three pairs of conserved cysteine residues, and the presence of a single added fucose residue on the EGF domain, but not of other hexoses or hexosamines. Efficient fucosylation with a single fucosylated EGF domain species can be detected by HPLC, strongly suggesting that correct folding of the nascent EGF peptide occurs in the yeast ER. This system appears to be useful for the systematic analysis of folding or secretion of EGF domain-containing proteins in the absence of native cell-specific factors. Previous to this study, both glycine residues of the EGF domain consensus sequence were shown to be important for correct disulfide pairing, leading to recognition by O-FucT-1 (Wang and Spellman 1998). Nonetheless, the CADASIL mutation in 5th EGF repeat of Notch3, in which the third cysteine in the repeat is replaced by serine, has been reported not to prevent O-fucosylation (Arboleda-Velasquez et al. 2005). Thus, it is not well understood if all three disulfide bonds in the EGF domain are necessary for O-fucosylation by O-FucT-1 or not, and their relative roles in protein folding and secretion remain unclear. To address this, we constructed mutant EGF domains that eliminated each of three disulfide bonds and expressed them in the engineered yeast system. Our finding that the 1-3 mutant is O-fucosylated is consistent with the results with the CADASIL Notch3 mutant protein and further, our results indicate that the EGF domain containing only the C2–C4 and C5–C6 disulfide linkages can be recognized by O-FucT-1. In contrast, the 2-4 mutant protein did not contain an O-fucose residue, despite the fact that equivalent amounts of the potential substrate protein were produced in cells. Thus, the C2–C4 disulfide linkage appears to be important for correct folding, which in turn leads to recognition by the O-FucT-1 enzyme.
The results of a comparison of EGF domains of different proteins suggested that domains that are linked to O-fucose do not always interact with AAL. In other work, binding with AAL has been used to detect an O-fucose structure (Ishikawa et al. 2005). Our results suggest that this method is useful but not sufficient, because AAL can bind the EGF domains of factor VII but not of factor IX. Furthermore, we found that in addition to differences in the mobility of O-fucosylated factor IX versus factor VII EGF domain on SDS–PAGE, the EGF domain from factor IX but not from factor VII serves as a substrate for human Manic Fringe in the yeast system containing no mammalian cell-specific factors. These data suggest that not only the amino acid sequence, as reported previously (Rampal et al. 2005), but also a three-dimensional structure of O-fucosylated EGF domain itself may contain information that is specifically recognized by Manic Fringe.
The construction of an O-fucosylation system in yeast also revealed a GDP-fucose transport activity in yeast. There have been several recent reports of broad specificity of NSTs (Berninsone et al. 2001; Segawa et al. 2002; Hong et al. 2000). In particular, Leishmania LPG2 NST can transport GDP-fucose and GDP-arabinose as well as GDP-mannose (Hong et al. 2000). Initially, we assumed that Vrg4p, which encodes GDP-mannose transporter (Dean et al. 1997), might function as a GDP-fucose transporter in yeast. To address this possibility, we tested the effect of a mutant allele of the gene that encodes GDP-mannose, vrg4-2 (Dean et al. 1997), in yeast cells equipped with the exogenous O-fucosylation system. Both the amount of O-fucose in the secreted EGF domain and in vitro activity of GDP-fucose transport in the vrg4-2 mutant strain were about half of that observed in cells without the vrg4-2 mutation. However, the introduction of wild-type VRG4 failed to recover in vivo O-fucosylation or in vitro GDP-fucose transport activity (data not shown). These results suggest that GDP-fucose transport activity may not be attributable to Vrg4p and thus, an as yet unidentified GDP-fucose transporter may be present in S. cerevisiae despite the fact that these yeasts do not exhibit endogenous fucosylation. This is similar to a previously observed phenomenon; namely, that S. cerevisiae has UDP-Gal transport activity, as in vivo galactosylation occurs in cells equipped with exogenous galactosylation machinery; in this case as well, the identity of the UDP-Gal transporter is not known (Roy et al. 1998).
Recently, an in vivo system for producing UDP-xylose was engineered in yeast by expressing the enzymes needed to convert UDP-glucose to UDP-xylose in this cell type (Oka and Jigami 2006). It is easy to envision, then, that it should soon be possible to use a similar strategy to produce mammalian O-glycans containing heterologous sugars, such as O-GalNAc, O-xylose, and O-glucose in yeast, in addition to the production of O-fucose-modified polypeptides as demonstrated here.
|
Materials and methods |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsFundingConflict of interest statementReferences |
|---|
Strains, culture conditions, and reagents – S. cerevisiae strains W303-1A (MATa leu2-3,112 his3-11,15 ade2-1 ura3-1 trp1-1 can1-100) and W303-1B (MAT
leu2-3,112 his3-11,15 ade2-1 ura3-1 trp1-1 can1-100) – were used as host cells for in vivo construction of an artificial O-glycosylation system. Other strains used are listed in Table I. Yeasts were grown in YPAD (1% yeast extract, 2% peptone, 2% glucose, 0.003% adenine) or YPADP (YPAD containing 0.1 M potassium phosphate, pH 6.5).
Construction of plasmids for in vivo O-glycosylation
The human GFR gene, which encodes a GDP-fucose trans- porter, was amplified from a cDNA library (human brain QUICK-Clone cDNA; Clontech, Palo Alto, CA) using the oligonucleotide primers 5'-AATGAGCTCATGAATAGGGC- CCCTCTGAAG-3' and 5'-ACTCTAGATCATTTACCCAAT- CTATTCATTTCAATATCAGTGTACACCCCCATGGCGCT- CTTC-3', which were designed to add a VSV-G tag coding sequence. The PCR fragment was digested with SacI and XbaI and cloned into SacI and XbaI-digested YEp352GAP expression vector (2 µ; URA3), which contains the TDH3 promoter and terminator. The resulting construct was designated YEp-hGFR. Additionally, BamHI-digested YEp-hGFR, a fragment containing the hGFR gene, and the TDH3 promoter/terminator was treated with KOD DNA polymerase (TOYOBO, Osaka, Japan) for gap-filling and repair and then ligated into the PvuII site of the integration vector pRS306 (URA3). The resulting construct was designated pRS-hGFR.
The human POFUT1 gene, which encodes O-FucT-1, was amplified from a human cDNA library (human brain QUICK-Clone cDNA; Clontech) using the primers 5'-AGAATTCATGGGCGCCGCCG-3' and 5'-GCTCCGG- CTCGAGTCAGAACTCGTCCCGCA-3'. The PCR fragment was digested with EcoRI and XhoI and cloned into EcoRI and XhoI-digested YEp352GAP. Subsequently, a gene with SpeI and BglII restriction sites was produced by replacing the section of the 5'-region of POFUT1 that contains the transmembrane domain with a PCR fragment amplified using the primers 5'-AGAATTCATGGGCGCCGCCG-3' and 5'-AGCCCGCGGGCATTGAGATCTGTACTAGTCCC GGGAGCGGCAGAAGCAGC-3', which is a FLAG tag. The FLAG tag from the yeast vector pESC-trp (Stratagene, La Jolla, CA) was introduced into the SpeI and BglII sites of POFUT1. The plasmid was then digested with BamHI and the resulting fragment, which contains both POFUT1 and the TDH3 promoter/terminator, was treated with polymerase (TOYOBO) for gap-filling and repair and then ligated into the PvuII site of the integration vector pRS305 (LEU2). The resulting construct was designated pRS-POFUT1.
Human manic fringe gene, which encodes β1,3-N-acetylglucosaminyltransferase, was amplified from a human cDNA library (human brain QUICK-Clone cDNA; Clontech) using the primers 5'-CCGAATTCATGCAGTGCCGG- CTCCCGCG-3' and 5'-GGGTCGACCTATTTACCCAATCTA- TTCATTTCAATATCAGTGTATCGGGCACCCAGCTGGG- GAC-3', which was designed to add a VSV-G tag coding sequence. The amplified fragment was digested with EcoRI and SalI and cloned into EcoRI and SalI-digested pRS-TDH, which contains the URA3 marker and the TDH3 promoter and terminator. The resulting fringe expression vector was designated pRS-mFLG.
The MURI, AtFX expression vector was constructed as follows. A PvuI fragment containing the TDH3 promoter/terminator and MUR1 was obtained from the yeast vector YEp-MUR1HA (Nakayama et al. 2003) and ligated into the PvuI site of the integration vector pRS303 (HIS3), generating pRS-MUR1. A BamHI fragment containing the TDH3 promoter and AtFX was obtained from the expression vector pYO-AtFXMyc (Nakayama et al. 2003), treated with polymerase and blunt-end ligated into the NaeI site of pRS-MUR1, generating pRS-MUR1/AtFX.
The human factor VII EGF-1 domain was used as an acceptor for O-FucT-1. The acceptor consisted of amino acid residues 105–147 of the mature factor VII protein (Wang et al. 1996) with an N-terminal 6xHistidine tag (AGHHHHHHVSDGDQCASSPCQNGGSCKDQLQSYICFCLPAFEGRNCETHKDD). The construct was produced by annealing primers designed to form oligonucleotide cassettes using standard molecular biology techniques and then cloned into the ApaI and KpnI-digested pCR2.1 (Invitrogen, Carlsbad, CA). The EGF domain coding sequence was then digested with NaeI and KpnI, and introduced into similarly digested pAFF2, which includes a -factor pre-pro sequence to direct secretion of the gene product by yeast. An expression cassette comprised of the -factor pre-pro region, 6xHis tag, and EGF domain-containing coding sequence under the control of the TDH3 promoter and terminator was isolated via BamHI digestion. The BamHI fragment was then introduced into plasmid pJJ246 (TRP1), generating pJJ-EGFVII.
The human factor IX EGF domain-1 construct was generated from pSYN-758 (Takara Bio Inc. Shiga, Japan), which contains the factor IX EGF domain-1 fused with a hexahistidine tag coding sequence at the 5'-terminus. NaeI and ApaI sites were added at the 5'- and 3'-termini, the plasmid was digested with EcoRI and HindIII, and the isolated fragment was introduced into EcoRI- and HindIII-digested pBluescript II (Stratagene, La Jolla, CA). This plasmid was then digested with NaeI and XhoI, and the coding sequence-containing fragment was introduced into the vector backbone fragment of NaeI and XhoI-digested pJJ-EGFVII, resulting in pJJ-EGFIX.
Plasmids useful for expression of C1A-C3A (1-3egfVII), C2A-C4A (2-4egfVII), and C5A-C6A (5-6egfVII) mutant domains were constructed as follows. Mutations were introduced into fragments of the coding sequence via PCR using the forward primers C1A-C3A-F (5'-TGATGG- TGATCAAGCTGCTTCTTCTCCATGTCAAAACGGTGGT- TCTGCTAAGGACCAATTGCAATCTTAC-3'), C2A-C4A-F (5'-TGATGGTGATCAATGTGCTTCTTCTCCAGCTCAAA- ACGGTGGTTCTTGTAAGGACCAATTGCAATCTTAC-3'), and C5A-C6A-F (5'-TGATGGTGATCAATGTGCTT- CTTCTCCATGTCAAAACGGTGGTTCTTGTAAGGACCA- ATTGCAATCTTAC-3') and the reverse primers C1A-C3A-R (5'-TTAGTCATCCTTATGAGTTTCACAGTTTCTACCTT- CGAAAGCTGGCAAACAGAAACAAATGTAAGATTGCA- ATTGG-3'), C2A-C4A-R (5'-TTAGTCATCCTTATGAGTT- TCACAGTTTCTACCTTCGAAAGCTGGCAAACAGAAA- GCAATGTAAGATTGCAATTGG-3'), and C5A-C6A-R (5'-TTAGTCATCCTTATGAGTTTCAGCGTTTCTACCTTCGA- AAGCTGGCAAAGCGAAACAAATGTAAGATTGCAATT- GG-3') for 1-3egfVII, 2-4egfVII, and 5-6egfVII, respectively. Next, the complete coding regions of 1-3egfVII, 2-4egfVII, and 5-6egfVII were amplified by PCR using each partial fragment as a template for the primers 5'-AAAAA- GCCGGCCATCACCATCACCATCACGTGTCTGATGGTG- ATCAATGT-3' and 5'-AAAGGTACCTTAGTCATCCTTATG- AGTTTC-3'. The amplified DNA fragments were digested with NaeI and KpnI, and inserted into similarly digested pAFF2. A DNA fragment containing the mutant EGF domain-encoding sequence, -factor pre-pro region, and TDH3 promoter and terminator was obtained by BamHI digestion. This fragment was inserted into the BamHI site of pRS306 (URA3) to yield pRS-1-3egfVII, pRS-2-4egfVII, and pRS-5-6egfVII.
Construction of yeast strains
The integration plasmids pRS-hGFR, pRS-POFUT1, pJJ-EGFVII, pJJ-EGFIX, and pRS-mFLG were digested with EcoRV and pRS-MUR1/AtFX was digested with NheI. The integration plasmids pRS-1-3egfVII, pRS-2-4egfVII, and pRS-5-6egfVII were digested with StuI. The linearized fragments were transformed into yeast using standard methods and resulted in the integration of the expression cassettes into the genomic URA3, LEU2, TRP1, or HIS3 locus of the strain W303-1A or W303-1B. Correct integration of each gene fragment was confirmed by PCR and expression was examined by immunoblotting. For each strain, engineered cells were selected on an appropriate nutritional drop-out medium after tetrad analysis.
Detection of recombinant proteins
Yeast cells of the appropriate strains were grown in 3 mL of YPAD medium for 12 h at 30°C and collected by centrifugation at 3000 x g for 5 min. The cells were then washed with 10 mM sodium azide and resuspended in 100 µL of lysis buffer (100 mM NaCl, 50 mM Tris–HCl, pH 7.5, 1 mM EDTA, 1 mM PMSF, and protease inhibitor cocktail; Roche Diagnostics, Mannheim, Germany). The cells were lysed by the addition of glass beads and vortexing for 30 s. The cell suspension was dissolved with 1% Triton X-100 and centrifuged at 5000 x g for 10 min. The supernatant was either mixed with an equal amount of SDS–PAGE sample buffer (100 mM Tris–HCl, pH 6.8, 2% SDS, 6% [v/v] 2-mercaptoethanol, and 10% [v/v] glycerol) and incubated for 1 h at room temperature, or used immediately in the immunoprecipitation assay.
For immunoprecipitation, ANTI-FLAG M2 Affinity Gel (Sigma Aldrich, St. Louis, MO) or VSV-G Antibody Agarose Immobilized (Bethyl Inc. Montgomery, TX) was added, and the suspension was shaken for 24 h at 4°C and then mixed with an electrophoresis sample buffer. The samples were analyzed by immunoblotting with one of the following antibodies: goat anti-VSV-G, rabbit anti-VSV-G, mouse anti-HA, mouse anti-Myc, or mouse anti-FLAG. Protein bands recognized by the primary antibodies were visualized using horseradish peroxidase-conjugated secondary antibodies and the ECL Plus system (GE Healthcare, Little Chalfont, Buckinghamshire, U K).
Purification of secreted EGF domains and lectin blotting
Cells were grown in 40 mL of the YPAD or YPADP medium at 30°C for the indicated period and harvested by centrifugation at 5000 x g for 10 min. The pH of the culture supernatant was adjusted to 8.0 with NaOH. Next, Ni-NTA agarose (QIAGEN GmbH, Hilden, Germany) was added to the supernatant and the suspension was shaken for 2 h at 4°C. The Ni-NTA agarose resin was then packed into a column and washed with a W1 buffer (50 mM NaH2PO4, 0.3 M NaCl, and 10 mM imidazole, pH 8.0). Protein was eluted with an E buffer (50 mM NaH2PO4, 0.3 M NaCl, and 250 mM imidazole, pH 8.0). The eluted protein was analyzed by immunoblotting using a peroxidase-conjugated anti-His5 antibody for detection (QIAGEN GmbH).
Lectin blotting was used to detect fucosylation of the heterologously expressed EGF domain. Following electrophoresis, proteins were transferred to a polyvinylidene difluoride membrane and then blocked with a blocking buffer (PBS buffer containing 3% BSA) for 1 h at room temperature. The membrane was then transferred to 1:1000 AAL lectin-biotin (Seikagaku Corporation, Tokyo, Japan) in the blocking buffer and incubated for 12 h at 4°C. Protein bands containing fucose were visualized using the horseradish peroxidase-conjugated avidin system, followed by treatment with ECL Plus reagents (GE Healthcare).
Purification of EGF domains from cell extracts
Yeast cells were grown in 500 mL of YPADP medium for 24 h at 30°C and collected by centrifugation at 3000 x g for 5 min. The cells were then washed with a W2 buffer (0.3 M NaCl, 50 mM Tris–HCl, pH 7.5) for four times and resuspended in 15 mL of lysis buffer B (8 M urea, 0.5% SDS, 0.3 M NaCl, 50 mM Tris–HCl, pH 7.5, and EDTA-free protease inhibitor cocktail from Roche Diagnostics). The cells were lysed by the addition of glass beads and vortexing for 3 min. The cell suspension was centrifuged at 10,000 x g for 10 min. Ni-NTA agarose (QIAGEN) was added to the supernatant of cell extract and the suspension was shaken for 2 h at 4°C. The Ni-NTA agarose resin was washed with a W1 buffer (50 mM NaH2PO4, 0.3 M NaCl, and 10 mM imidazole, pH 8.0). Peptide was eluted with the E buffer (50 mM NaH2PO4, 0.3 M NaCl, and 250 mM imidazole, pH 8.0).
HPLC
HPLC was performed using a Shimadzu CLASS-VP system (Shimadzu Corporation, Kyoto, Japan) with a C-4 butyl Cosmosil column (4.6 x 150 mm; nacalai tesque; Kyoto, Japan) at a flow rate of 1 mL/min. Buffer A was 0.1% aqueous trifluoroacetic acid (TFA) and buffer B was acetonitrile (CH3CN) containing 0.1% TFA. After injection, the concentration of buffer B was increased from 0% to 80% for 50 min. The column effluent was monitored at 215 nm for the elution of proteins.
MALDI-ToF MS
The HPLC-purified EGF domain sample was dissolved in water and mixed with -cyano-4-hydroxycinnamic acid saturated solution (0.1% TFA and 50% CH3CN) as a matrix and applied to a target board. The average mass was determined in positive mode using an Ettan MALDI-ToF/Pro mass spectrometer (GE Healthcare). We used average masses to calculate theoretical masses.
Analysis of the GlcNAc-Fuc on the secreted EGF domain
Monosaccharides from the EGF domain sample were prepared by acid hydrolysis. To do this, the HPLC-purified protein sample was incubated in 2N TFA at 100°C for 5 h and dried. Next, the hydrolysates were acetylated using an acetylating agent (Seikagaku Corporation) and labeled with fluorescent ABEE using an ABEE labeling kit (Seikagaku Corporation) according to the manufacturer's protocol. The ABEE-labeled monosaccharides were analyzed by HPLC using a Honenpak C18 column (4.6x 75 mm; Seikagaku Corporation) at a flow rate of 1 mL/min with 0.2 M potassium borate buffer (pH 9.0) containing 6% CH3CN at 45°C. The elution time of the monosaccharides was determined using the ABEE-labeled standard (Yasuno et al. 1999).
|
Funding |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsFundingConflict of interest statementReferences |
|---|
The New Energy and Industrial Technology Development Organization of Japan.
|
Conflict of interest statement |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsFundingConflict of interest statementReferences |
|---|
None declared.
|
Acknowledgements |
|---|
We thank Dr. N. Dean for providing yeast strains used in the study.
|
Abbreviations |
|---|
AAL, Aleuria auramtia lectin; ABEE, p-aminobenzoic acid ethylester; EGF, epidermal growth factor; ER, endoplasmic reticulum; GFR, GDP-fucose transporter; GlcNAc, N-acetylglucosamine; HexNAc, N-acetylhexosamine; hGFR, human GFR; MALDI-ToF, matrix-assisted laser desorption ionization-time of flight; NST, nucleotide sugar transporter; O-FucT-1, protein O-fucosyltransferase 1
|
References |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsFundingConflict of interest statementReferences |
|---|
Arboleda-Velasquez JF, Rampal R, Fung E, Darland DC, Liu M, Martinez MC, Donahue CP, Navarro-Gonzalez MF, Libby P, D’Amore PA, et al. CADASIL mutations impair Notch3 glycosylation by Fringe. Hum Mol Genet (2005) 14:1631–1639.
Berninsone P, Hwang HY, Zemtseva I, Horvitz HR, Hirschberg CB. SQV-7, a protein involved in Caenorhabditis elegans epithelial invagination and early embryogenesis, transports UDP-glucuronic acid, UDP-N-acetylgalactosamine, and UDP-galactose. Proc Natl Acad Sci USA (2001) 98:3738–3743.
Bjoern S, Foster DC, Thim L, Wiberg FC, Christensen M, Komiyama Y, Pedersen AH, Kisiel W. Human plasma and recombinant factor VII. Characterization of O-glycosylations at serine residues 52 and 60 and effects of site-directed mutagenesis of serine 52 to alanine. J Biol Chem (1991) 266:11051–11057.
Bruckner K, Perez L, Clausen H, Cohen S. Glycosyltransferase activity of Fringe modulates Notch-Delta interactions. Nature (2000) 406:411–415.[CrossRef][Medline]
Chiba Y, Suzuki M, Yoshida S, Yoshida A, Ikenaga H, Takeuchi M, Jigami Y, Ichishima E. Production of human compatible high mannose-type (Man5GlcNAc2) sugar chains in Saccharomyces cerevisiae. J Biol Chem (1998) 273:26298–26304.
Choi BK, Bobrowicz P, Davidson RC, Hamilton SR, Kung DH, Li H, Miele RG, Nett JH, Wildt S, Gerngross TU. Use of combinatorial genetic libraries to humanize N-linked glycosylation in the yeast Pichia pastoris. Proc Natl Acad Sci USA (2003) 100:5022–5027.
Dean N, Zhang YB, Poster JB. The VRG4 gene is required for GDP-mannose transport into the lumen of the Golgi in the yeast, Saccharomyces cerevisiae. J Biol Chem (1997) 272:31908–31914.
Hallgren P, Lundblad A, Svensson S. A new type of carbohydrate-protein linkage in a glycopeptide from normal human urine. J Biol Chem (1975) 250:5312–5314.
Harris RJ, Spellman MW. O-Linked fucose and other post-translational modifications unique to EGF modules. Glycobiology (1993) 3:219–224.
Helenius A, Aebi M. Intracellular functions of N-linked glycans. Science (2001) 291:2364–2369.
Hong K, Ma D, Beverley SM, Turco SJ. The Leishmania GDP-mannose transporter is an autonomous, multi-specific, hexameric complex of LPG2 subunits. Biochemistry (2000) 39:2013–2022.[CrossRef][Web of Science][Medline]
Ishikawa HO, Higashi S, Ayukawa T, Sasamura T, Kitagawa M, Harigaya K, Aoki K, Ishida N, Sanai Y, Matsuno K. Notch deficiency implicated in the pathogenesis of congenital disorder of glycosylation IIc. Proc Natl Acad Sci USA (2005) 102:18532–18537.
Kentzer EJ, Buko A, Menon G, Sarin VK. Carbohydrate composition and presence of a fucose-protein linkage in recombinant human pro-urokinase. Biochem Biophys Res Commun (1990) 171:401–406.[CrossRef][Web of Science][Medline]
Kui Wong N, Easton RL, Panico M, Sutton-Smith M, Morrison JC, Lattanzio FA, Morris HR, Clark GF, Dell A, Patankar MS. Characterization of the oligosaccharides associated with the human ovarian tumor marker CA125. J Biol Chem (2003) 278:28619–28634.
Li H, Sethuraman N, Stadheim TA, Zha D, Prinz B, Ballew N, Bobrowicz P, Choi BK, Cook WJ, Cukan M, et al. Optimization of humanized IgGs in glycoengineered Pichia pastoris. Nat Biotechnol (2006) 24:210–215.[CrossRef][Web of Science][Medline]
Lubke T, Marquardt T, Etzioni A, Hartmann E, von Figura K, Korner C. Complementation cloning identifies CDG-IIc, a new type of congenital disorders of glycosylation, as a GDP-fucose transporter deficiency. Nat Genet (2001) 28:73–76.[CrossRef][Web of Science][Medline]
Luhn K, Wild MK, Eckhardt M, Gerardy-Schahn R, Vestweber D. The gene defective in leukocyte adhesion deficiency II encodes a putative GDP-fucose transporter. Nat Genet (2001) 28:69–72.[CrossRef][Web of Science][Medline]
Moloney DJ, Panin VM, Johnston SH, Chen J, Shao L, Wilson R, Wang Y, Stanley P, Irvine KD, Haltiwanger RS, et al. Fringe is a glycosyltransferase that modifies Notch. Nature (2000) 406:369–375.[CrossRef][Medline]
Nakajima T, Ballou CE. Characterization of the carbohydrate fragments obtained from Saccharomyces cerevisiae mannan by alkaline degradation. J Biol Chem (1974a) 249:7679–7684.
Nakajima T, Ballou CE. Structure of the linkage region between the polysaccharide and protein parts of Saccharomyces cerevisiae mannan. J Biol Chem (1974b) 249:7685–7694.
Nakayama K, Maeda Y, Jigami Y. Interaction of GDP-4-keto-6-deoxymannose-3,5-epimerase-4-reductase with GDP-mannose-4,6-dehydratase stabilizes the enzyme activity for formation of GDP-fucose from GDP-mannose. Glycobiology (2003) 13:673–680.
Nishimura H, Takao T, Hase S, Shimonishi Y, Iwanaga S. Human factor IX has a tetrasaccharide O-glycosidically linked to serine 61 through the fucose residue. J Biol Chem (1992) 267:17520–17525.
Oka T, Jigami Y. Reconstruction of de novo pathway for synthesis of UDP-glucuronic acid and UDP-xylose from intrinsic UDP-glucose in Saccharomyces cerevisiae. FEBS J (2006) 273:2645–2657.[CrossRef][Medline]
Okajima T, Irvine KD. Regulation of notch signaling by O-linked fucose. Cell (2002) 111:893–904.[CrossRef][Web of Science][Medline]
Panin VM, Shao L, Lei L, Moloney DJ, Irvine KD, Haltiwanger RS. Notch ligands are substrates for protein O-fucosyltransferase-1 and Fringe. J Biol Chem (2002) 277:29945–29952.
Rabbani SA, Mazar AP, Bernier SM, Haq M, Bolivar I, Henkin J, Goltzman D. Structural requirements for the growth factor activity of the amino-terminal domain of urokinase. J Biol Chem (1992) 267:14151–14156.
Rampal R, Li AS, Moloney DJ, Georgiou SA, Luther KB, Nita-Lazar A, Haltiwanger RS. Lunatic fringe, manic fringe, and radical fringe recognize similar specificity determinants in O-fucosylated epidermal growth factor-like repeats. J Biol Chem (2005) 280:42454–42463.
Roy SK, Yoko-o T, Ikenaga H, Jigami Y. Functional evidence for UDP-galactose transporter in Saccharomyces cerevisiae through the in vivo galactosylation and in vitro transport assay. J Biol Chem (1998) 273:2583–2590.
Segawa H, Kawakita M, Ishida N. Human and Drosophila UDP-galactose transporters transport UDP-N-acetylgalactosamine in addition to UDP-galactose. Eur J Biochem (2002) 269:128–138.[Web of Science][Medline]
Shao L, Moloney DJ, Haltiwanger R. Fringe modifies O-fucose on mouse Notch1 at epidermal growth factor-like repeats within the ligand-binding site and the Abruptex region. J Biol Chem (2003) 278:7775–7782.
Shi S, Stanley P. Protein O-fucosyltransferase 1 is an essential component of Notch signaling pathways. Proc Natl Acad Sci USA (2003) 100:5234–5239.
Wang Y, Lee GF, Kelley RF, Spellman MW. Identification of a GDP-L-fucose:polypeptide fucosyltransferase and enzymatic addition of O-linked fucose to EGF domains. Glycobiology (1996) 6:837–842.
Wang Y, Shao L, Shi S, Harris RJ, Spellman MW, Stanley P, Haltiwanger RS. Modification of epidermal growth factor-like repeats with O-fucose. Molecular cloning and expression of a novel GDP-fucose protein O-fucosyltransferase. J Biol Chem (2001) 276:40338–40345.
Wang Y, Spellman MW. Purification and characterization of a GDP-fucose:polypeptide fucosyltransferase from Chinese hamster ovary cells. J Biol Chem (1998) 273:8112–8118.
Yasuno S, Kokubo K, Kamei M. New method for determining the sugar composition of glycoproteins, glycolipids, and oligosaccharides by high-performance liquid chromatography. Biosci Biotechnol Biochem (1999) 63:1353–1359.[CrossRef][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||