Engineering intracellular CMP-sialic acid metabolism into insect cells and methods to enhance its generation

Previous studies have reported that insect cell lines lack the capacity to generate endogenously the nucleotide sugar, CMP-Neu5Ac, required for sialylation of glycoconjugates. In this study, the biosynthesis of this activated form of sialic acid completely from endogenous metabolites is demonstrated for the first time in insect cells by expressing the mammalian genes required for the multistep conversion of endogenous UDP-GlcNAc to CMP-Neu5Ac. The genes for UDP-GlcNAc-2-epimerase/ManNAc kinase (EK), sialic acid 9-phosphate synthase (SAS), and CMP-sialic acid synthetase (CSAS) were coexpressed in insect cells using baculovirus expression vectors, but the CMP-Neu5Ac and precursor Neu5Ac levels synthesized were found to be lower than those achieved with ManNAc supplementation due to feedback inhibition of the EK enzyme by CMP-Neu5Ac. When sialuria-like mutant EK genes, in which the site for feedback regulation has been mutated, were used, CMP-Neu5Ac was synthesized at levels more than 4 times higher than that achieved with the wild-type EK and 2.5 times higher than that achieved with ManNAc feeding. Addition of N-acetylglucosamine (GlcNAc), a precursor for UDP-GlcNAc, to the media increased the levels of CMP-Neu5Ac even more to a level 7.5 times higher than that achieved with ManNAc supplementation, creating a bottleneck in the conversion of Neu5Ac to CMP-Neu5Ac at higher levels of UDP-GlcNAc. The present study provides a useful biochemical strategy to synthesize and enhance the levels of the sialylation donor molecule, CMP-Neu5Ac, a critical limiting substrate for the generation of complex glycoproteins in insect cells and other cell culture systems.     

Nucleotide sugar biosynthesis occurs in the glycosomes of procyclic and bloodstream form Trypanosoma brucei

In Trypanosoma brucei, there are fourteen enzymatic biotransformations that collectively convert glucose into five essential nucleotide sugars: UDP-Glc, UDP-Gal, UDP-GlcNAc, GDP-Man and GDP-Fuc. These biotransformations are catalyzed by thirteen discrete enzymes, five of which possess putative peroxisome targeting sequences. Published experimental analyses using immunofluorescence microscopy and/or digitonin latency and/or subcellular fractionation and/or organelle proteomics have localized eight and six of these enzymes to the glycosomes of bloodstream form and procyclic form T. brucei, respectively. Here we increase these glycosome localizations to eleven in both lifecycle stages while noting that one, phospho-N-acetylglucosamine mutase, also localizes to the cytoplasm. In the course of these studies, the heterogeneity of glycosome contents was also noted. These data suggest that, unlike other eukaryotes, all of nucleotide sugar biosynthesis in T. brucei is compartmentalized to the glycosomes in both lifecycle stages. The implications are discussed.     

Mechanism of UDP-sugar transport into intracellular vesicles. Occurrence of UDP-GlcNAc/UDP and UDP-Gal/UDP antiports

The mechanism of translocation of UDP-GlcNAc, UDP-Gal and UDP-Glc into intracellular vesicles has been studied using thymocytes whose plasma membranes have been permeabilized with isotonic ammonium chloride. It has been previously shown that the intracellular vesicles have specific carriers for UDP-GlcNAc and UDP-Gal. We now report that the translocation of these two sugar nucleotides occurs via UDP-GlcNAc/UDP and UDP-Gal/UDP antiports. The entry of UDP-GlcNAc or UDP-Gal into vesicles was specifically dependent on the exit of UDP from these vesicles. In contrast, no antiport mechanism has been recovered with UDP-Glc for which no transport and accumulation into intracellular vesicles were observed.     

Influence of intracellular nucleotide and nucleotide sugar contents on recombinant interferon-gamma glycosylation during batch and fed-batch cultures of CHO cells

Both the macroheterogeneity of recombinant human IFN-gamma produced by CHO cells and intracellular levels of nucleotides and sugar nucleotides, have been characterized during batch and fed-batch cultures carried out in different media. Whereas PF-BDM medium was capable to maintain a high percentage of the doubly- glycosylated glycoforms all over the process, mono-glycosylated and non-glycosylated forms increased during the batch culture using SF-RPMI medium. Intracellular level of UTP was higher in PF-BDM all over the batch culture compared to the SF-RPMI process. UDP-Gal accumulated only during the culture performed in PF-BDM medium, probably as a consequence of the reduced UDP-Glc synthesis flux in SF-RPMI medium. When the recombinant CHO cells were cultivated in fed-batch mode, the UTP level remained at a relatively high value in serum-containing RPMI and its titer increased during the fed-phase indicating an excess of biosynthesis. Besides, an accumulation of UDP-Gal occurred as well. Those results all together indicate that UTP and UDP-Glc syntheses in CHO cells cultivated in SF-RPMI medium in batch process, could be limiting during the glycosylation processes of the recombinant IFN-gamma. At last, the determination of the energetic status of the cells over the three studied processes suggested that a relationship between the adenylate energy charge and the glycosylation macroheterogeneity of the recombinant IFN-gamma may exist.     

Uridine sugar nucleotides modified in their nucleoside moieties and their effect on lipid-linked saccharide formationin vitro

Uridine diphospho glucose (UDP-Glc) and uridine diphospho N-acetylglucosamine (UDP-GlcNAc), modified in the uridine moiety by either periodate oxidation of the ribose ring or substitution at position 5 of the uracil ring with fluorine, have been tested as potential inhibitors of glucosyl monophosphoryl dolichol (Glc-P-Dol) or N,N-diacetylchitobiosyl pyrophosphoryl dolichol [GlcNAc)2-PP-Dol) assembly in chick embryo cell membranes. The periodate oxidised sugar nucleotides inhibited glycosyl transfer from their respective natural counterparts by 50% at 230 micron periodate oxidised UDP-Glc and 70 micron periodate oxidised UDP-GlcNAc respectively. Inhibition in both cases was irreversible and addition of exogenous Dol-P stimulated only the residual non-inhibited reaction. Periodate oxidised UDP-GlcNAc preferentially inhibited the transfer of GlcNAc to GlcNac-PP-Dol. The sugar nucleotide containing 5-fluorouridine were, on the other hand, alternative substrates for Glc-P-Dol or (GlcNAc)2-PP-Dol synthesis. FUDP-Glc was a good substrate for Glc-P-Dol formation; having Km and Vmax values equal to those of UDP-Glc, whereas FUDP-GlcNAc was a less efficient substrate for the formation of (GlcNAc)2-PP-Dol; having Km and Vmax values one half and one third respectively of those of UDP-GlcNAc.     

Membrane transport of sugar donors to the glycosylation sites

The assembly of N-linked glycoproteins in eukaryotic cells begins with the segregation of these molecules within the lumen of intracellular vesicles. Since the sugar nucleotides are cytoplasmic molecules, translocation of the sugar moiety across the membrane appears as a crucial event in the glycoprotein synthesis. This N-glycosylation process occurs in two different cytological sites: in the rough endoplasmic reticulum, the stepwise synthesis of a large lipid-linked oligosaccharide takes place, as well as its transfer to protein; then after trimming the immature glycoprotein is further elongated in the Golgi apparatus. In this paper, a brief review will be given of the present knowledge on the sugar donor transport across the membrane barrier to the glycosylation site. Based upon the transmembrane orientation of oligosaccharide lipid intermediates and on the localization of the glycosyltransferase active sites, the different processes required to translocate the sugar moieties during the preassembly of the dolichyl-pyrophosphate-oligosaccharides will be examined. Combining the different results, obtained in several laboratories, it is suggested that the Man5-GlcNAc2-lipid is synthesized on the cytoplasmic side directly from the sugar-nucleotides and then translocated to the lumenal face where the Glc3-Man9-GlcNAc2-lipid is completed using Man-P-Dol and Glc-P-Dol as transmembrane carriers of these sugars. Concerning the elongation process leading to assembly of the antennae of N-acetyllactosamine type oligosaccharides, specific carriers for sugar nucleotides have been described as Golgi markers. Several authors have characterized such carriers for UDP-Gal, GDP-Fuc, CMP-NeuAc, UDP-GlcNAc and UDP-Glc using microsomal vesicles and similar results have been obtained in our laboratory using plasma membrane permeabilized cells. This carrier-mediated process leads to the formation of an intralumenal pool whose biological significance will be discussed. The translocation process of sugar donors occurring in the rough endoplasmic reticulum via lipid intermediates as well as in the Golgi apparatus via specific carriers would represent a regulation step based on the availability of the substrates for the glycosylation.     

Effect of anion-specific inhibitors on the utilization of sugar nucleotides for N-linked carbohydrate unit assembly by thyroid endoplasmic reticulum vesicles

The effect of anion-specific inhibitors on the utilization of the sugar nucleotides (UDP-glucose, GDP-mannose, and UDP-N-acetylglucosamine) required for the formation of the oligosaccharide-lipid involved in N-glycosylation has been studied in intact endoplasmic reticulum (ER) vesicles from thyroid. Of the reagents tested, the nonpenetrating probe DIDS (4,4'-diisothiocyano-2,2'-stilbenedisulfonic acid) and its dihydro derivative (H2DIDS) were the most effective, causing a pronounced impairment in the synthesis from UDP-Glc of dolichyl phosphate (Dol-P) glucose (50% reduction at 60 microM DIDS) and in the incorporation of glucose into oligosaccharide-lipid and N-glycosylated protein; in contrast, no inhibition was observed in the formation from UDP-Glc of a glycogen-like proteoglucan. The specificity of the DIDS effect was indicated by the finding that methyl isothiocyanate, a nonanionic amino-reactive agent, demonstrated negligible inhibition. While DIDS also effected a block in the formation of Dol-P-P-GlcNAc from UDP-GlcNAc, no impairment in the utilization of GDP-Man for Dol-P-Man synthesis was observed. Since the DIDS inhibition of UDP-Glc and UDP-GlcNAc utilization was maintained after disruption of the ER vesicles with Triton, even when the incubations were supplemented with Dol-P, it appears that this reagent does not interact with sugar nucleotide translocator proteins but rather with the cytoplasmically oriented anion binding sites of glycosyltransferases (UDP-Glc- and UDP-GlcNAc:Dol-P glucosyl- and GlcNAc-1-P transferases). This is consistent with the protease sensitivity of these enzymes in the intact ER vesicles. Incubation of the vesicles with tritiated H2DIDS (8 microM) introduced radioactivity into membrane polypeptides with molecular weights of about 52,000 and 31,000 as observed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, suggesting that this inhibitor may prove useful as an affinity label in further studies of some of the glycosyltransferases involved in the synthesis of lipid-monosaccharide intermediates.     

Cloning and expression of human sialic acid pathway genes to generate CMP-sialic acids in insect cells

The addition of sialic acid residues to glycoproteins can affect important protein properties including biological activity and in vivo circulatory half-life. For sialylation to occur, the donor sugar nucleotide cytidine monophospho-sialic acid (CMP-SA) must be generated and enzymatically transferred to an acceptor oligosaccharide. However, examination of insect cells grown in serum-free medium revealed negligible native levels of the most common sialic acid nucleotide, CMP-N-acetylneuraminic acid (CMP-Neu5Ac). To increase substrate levels, the enzymes of the metabolic pathway for CMP-SA synthesis have been engineered into insect cells using the baculovirus expression system. In this study, a human CMP-sialic acid synthase cDNA was identified and found to encode a protein with 94% identity to the murine homologue. The human CMP-sialic acid synthase (Cmp-Sas) is ubiquitously expressed in human cells from multiple tissues. When expressed in insect cells using the baculovirus vector, the encoded protein is functional and localizes to the nucleus as in mammalian cells. In addition, co-expression of Cmp-Sas with the recently cloned sialic acid phosphate synthase with N-acetylmannosamine feeding yields intracellular CMP-Neu5Ac levels 30 times higher than those observed in unsupplemented CHO cells. The absence of any one of these three components abolishes CMP-Neu5Ac production in vivo. However, when N-acetylmannosamine feeding is omitted, the sugar nucleotide form of deaminated Neu5Ac, CMP-2-keto-3-deoxy-D-glycero-D-galacto-nononic acid (CMP-KDN), is produced instead, indicating that alternative sialic acid glycoforms may eventually be possible in insect cells. The human CMP-SAS enzyme is also capable of CMP-N-glycolylneuraminic acid (CMP-Neu5Gc) synthesis when provided with the proper substrate. Engineering the CMP-SA metabolic pathway may be beneficial in various cell lines in which CMP-Neu5Ac production limits sialylation of glycoproteins or other glycans.     

Engineering sialic acid synthetic ability into insect cells: identifying metabolic bottlenecks and devising strategies to overcome them

Previous studies have indicated negligible levels of both sialylation and the precursor N-acetylneuraminic acid (Neu5Ac) in a number of insect cell lines grown in serum-free medium. The overexpression of the human sialic acid 9-phosphate synthase (SAS) in combination with N-acetylmannosamine (ManNAc) feeding has been shown to overcome this limitation. In this study we evaluated the potential bottlenecks in the sialic acid synthesis pathway in a Spodoptera frugiperda (Sf9) insect cell line and devised strategies to overcome them by overexpression of the enzymatic pathway enzymes combined with appropriate substrate feeding. Coexpression of SAS and UDP-GlcNAc 2-epimerase/ManNAc kinase, the bifunctional enzyme initiating sialic acid biosynthesis in mammals, resulted in Neu5Ac synthesis without use of any external media supplementation to demonstrate that Neu5Ac could be generated intracellularly in Sf9 cells using natural metabolic precursors. N-Acetylglucosamine (GlcNAc) feeding in combination with this coexpression resulted in much higher levels of Neu5Ac compared to levels obtained with ManNAc feeding with SAS expression alone. The lower Neu5Ac levels obtained with ManNAc feeding suggested limitations in the transport and phosphorylation of ManNAc. The bottleneck in phosphorylation was likely due to utilization of GlcNAc kinase for phosphorylation of ManNAc in insect cells and was overcome by expression of ManNAc kinase. The transport limitation was addressed by the addition of tetra-O-acetylated ManNAc, which is easily taken up by the cells. An alternative sialic acid, 2-keto-3-deoxy-D-glycero-D-galacto-nononic acid (KDN), could also be generated in insect cells, suggesting the potential for controlling not only the production of sialic acids but also the type of sialic acid generated. The levels of KDN could be increased with virtually no Neu5Ac generation when Sf9 cells were fed excess GlcNAc. The results of these studies may be used to enhance the sialylation of target glycoproteins in insect and other eukaryotic expression systems.     

Engineering nucleotide sugar synthesis pathways for independent and simultaneous modulation of N-glycan galactosylation and fucosylation in CHO cells

Glycosylation of recombinant therapeutics like monoclonal antibodies (mAbs) is a critical quality attribute. N-glycans in mAbs are known to affect various effector functions, and thereby therapeutic use of such glycoproteins can depend on a particular glycoform profile to achieve desired efficacy. However, there are currently limited options for modulating the glycoform profile, which depend mainly on over-expression or knock-out of glycosyltransferase enzymes that can introduce or eliminate specific glycans but do not allow predictable glycoform modulation over a range of values. In this study, we demonstrate the ability to predictably modulate the glycoform profile of recombinant IgG. Using CRISPR/Cas9, we have engineered nucleotide sugar synthesis pathways in CHO cells expressing recombinant IgG for combinatorial modulation of galactosylation and fucosylation. Knocking out the enzymes UDP-galactose 4′-epimerase (Gale) and GDP-L-fucose synthase (Fx) resulted in ablation of de novo synthesis of UDP-Gal and GDP-Fuc. With Gale knock-out, the array of N-glycans on recombinantly expressed IgG is narrowed to agalactosylated glycans, mainly A2F glycan (89%). In the Gale and Fx double knock-out cell line, agalactosylated and afucosylated A2 glycan is predominant (88%). In the double knock-out cell line, galactosylation and fucosylation was entirely dependent on the salvage pathway, which allowed for modulation of UDP-Gal and GDP-Fuc synthesis and intracellular nucleotide sugar availability by controlling the availability of extracellular galactose and fucose. We demonstrate that the glycoform profile of recombinant IgG can be modulated from containing predominantly agalactosylated and afucosylated glycans to up to 42% and 96% galactosylation and fucosylation, respectively, by extracellular feeding of sugars in a dose-dependent manner. By simply varying the availability of extracellular galactose and/or fucose, galactosylation and fucosylation levels can be simultaneously and independently modulated. In addition to achieving the production of tailored glycoforms, this engineered CHO host platform can cater to the rapid synthesis of variably glycoengineered proteins for evaluation of biological activity.     

Biochemical characterization of UDP-GlcNAc/Glc 4-epimerase from Escherichia coli O86:B7

The O-antigen of lipopolysaccharide in Gram-negative bacteria plays an important role in bacterium-host interactions. Escherichia coli O86:B7 O-unit contains five sugar residues: one fucose (Fuc) and two each of N-acetylgalactosamine (GalNAc) and galactose (Gal). The entire O-antigen gene cluster was previously sequenced: orf1 was assigned the gne gene for the biosynthesis of UDP-GalNAc. To confirm this annotation, overexpression, purification, and biochemical characterization of Gne were performed. By using capillary electrophoresis, we showed that Gne can catalyze the interconversion of both UDP-GlcNAc/GalNAc and UDP-Glc/Gal almost equally well. The Km values of Gne for UDP-Glc, UDP-Gal, UDP-GlcNAc, and UDP-GalNAc are 370, 295, 323, and 373 microM, respectively. The comparison of kinetic parameters of Gne from Escherichia coli O86:B7 to those of other characterized UDP-GlcNAc/Glc 4-epimerases indicated that it has relaxed specificity toward the four substrates, the first characterized enzyme to have this activity in the O-antigen biosynthesis. Moreover, the calculated kcat/Km values for UDP-GalNAc and UDP-Gal are approximately 2-4 times higher than those for UDP-GlcNAc and UDP-Glc, suggesting that Gne is slightly more efficient for the epimerization of UDP-GalNAc and UDP-Gal. One mutation (S306Y) resulted in a loss of epimerase activity for non-acetylated substrates by about 5-fold but totally abolished the activity for N-acetylated substrates, indicating that residue S306 plays an important role in the determination of substrate specificity.     

Thin-layer chromatographic method for the determination of glycosyltransferase activity

To survey glycosyltransferase activities and specificities we have developed a TLC method to separate various nucleotide sugars from both high- and low-molecular-weight sugar acceptors. Here, we report details of the procedure and its application for galactosyltransferase and fucosyltransferase detected in mouse spermatogenic cells. The assay method involves sample separation using polyethyleneimine cellulose plastic-backed thin-layer plates, developed in sodium phosphate buffer for 30 min. Nucleotide sugars, including UDP-Gal, GDP-Fuc, CMP-NeuNAc, and GDP-Man, remain at the origin, while both high- and low-molecular-weight sugar acceptors migrate within 2 cm of the solvent front. Assays for galactosyltransferase and fucosyltransferase are linear with time and yield results comparable to other methods such as gel permeation chromatography and micropartitioning filtration. The TLC protocol should be useful for determinations of many different glycosyltransferases.     

Intralumenal pool and transport of CMP-N-acetylneuraminic acid, GDP-fucose and UDP-galactose. Study with plasma-membrane-permeabilized mouse thymocytes.

Treatment with NH4Cl of mouse thymocytes renders their plasma membrane permeable to sugar nucleotides both inwards and outwards. Using this model, we studied the entry and utilization of CMP-NeuAc, GDP-Fuc and UDP-Gal into intracellular vesicles in situ. It is shown that CMP-NeuAc and GDP-Fuc enter the vesicles in a manner indicating a carrier-mediated transport (substrate saturation curve, inhibition by substrate analogues, temperature dependence) and are entrapped in their uncleaved form. This leads to the formation of an intralumenal pool of these precursors which can be further utilized by the sialyltransferases and fucosyltransferases. The occurrence of an endogenous pool of CMP-NeuAc and GDP-Fuc is demonstrated by the fact that, when the vesicles are disrupted by detergent, the release of the endogenous sugar nucleotides causes an isotopic dilution of the labelled precursors added to measure the glycosyltransferase activities. In contrast, no accumulation of UDP-Gal has been detected, suggesting that transport and transfer reaction are simultaneous events. However, experiments with UDP 2',3'-dialdehyde indicate that UDP-Gal is not transported through the membrane by galactosyltransferase action but by a distinct carrier molecule.     

Accumulation of UDP-GlcNAc into intracellular vesicles and occurrence of a carrier-mediated transport. Study with plasma-membrane-permeabilized mouse thymocytes

Metabolic labelling of mouse thymocytes with radioactive mannose or glucosamine leads to the formation of labelled GDP-Man and UDP-GlcNAc. Using isotonic ammonium chloride treatment which renders the plasma membrane of thymocytes permeable to sugar nucleotides, we demonstrate that, in contrast to GDP-Man, a pool of UDP-GlcNAc remains associated with the cells after plasma membrane permeabilization. These observations are confirmed in experiments in which permeabilized thymocytes are incubated with exogenous labelled GDP-Man and UDP-GlcNAc, and we show that only UDP-GlcNAc is accumulated into sealed intracellular vesicles. This accumulation is a saturable process which can be inhibited by UDP, demonstrating the occurrence of a specific carrier. This transport mechanism can be blocked by covalent attachment of a non-permeant inhibitor UDP-dialdehyde without affecting the N-acetylglucosaminyltransferase itself. The fact that this carrier-mediated transport is not inhibited by tunicamycin indicates that this translocation process of UDP-GlcNAc does not involve lipid intermediates.     

A single bifunctional UDP-GlcNAc/Glc 4-epimerase supports the synthesis of three cell surface glycoconjugates in Campylobacter jejuni

The major cell-surface carbohydrates (lipooligosaccharide, capsule, and glycoprotein N-linked heptasaccharide) of Campylobacter jejuni NCTC 11168 contain Gal and/or GalNAc residues. GalE is the sole annotated UDP-glucose 4-epimerase in this bacterium. The presence of GalNAc residues in these carbohydrates suggested that GalE might be a UDP-GlcNAc 4-epimerase. GalE was shown to epimerize UDP-Glc and UDP-GlcNAc in coupled assays with C. jejuni glycosyltransferases and in sugar nucleotide epimerization equilibria studies. Thus, GalE possesses UDP-GlcNAc 4-epimerase activity and was renamed Gne. The Km(app) values of a purified MalE-Gne fusion protein for UDP-GlcNAc and UDP-GalNAc are 1087 and 1070 microm, whereas those for UDP-Glc and UDP-Gal are 780 and 784 microm. The kcat and kcat/Km(app) values were three to four times higher for UDP-GalNAc and UDP-Gal than for UDP-GlcNAc and UDP-Glc. The comparison of the kinetic parameters of MalE-Gne to those of other characterized bacterial UDP-GlcNAc 4-epimerases indicated that Gne is a bifunctional UDP-GlcNAc/Glc 4-epimerase. The UDP sugar-binding site of Gne was modeled by using the structure of the UDP-GlcNAc 4-epimerase WbpP from Pseudomonas aeruginosa. Small differences were noted, and these may explain the bifunctional character of the C. jejuni Gne. In a gne mutant of C. jejuni, the lipooligosaccharide was shown by capillary electrophoresis-mass spectrometry to be truncated by at least five sugars. Furthermore, both the glycoprotein N-linked heptasaccharide and capsule were no longer detectable by high resolution magic angle spinning NMR. These data indicate that Gne is the enzyme providing Gal and GalNAc residues with the synthesis of all three cell-surface carbohydrates in C. jejuni NCTC 11168.     

Biosynthesis of GlcNAc-rich N- and O-glycans in the Golgi apparatus does not require the nucleotide sugar transporter SLC35A3

Nucleotide sugar transporters, encoded by the SLC35 gene family, deliver nucleotide sugars throughout the cell for various glycosyltransferase-catalyzed glycosylation reactions. GlcNAc, in the form of UDP-GlcNAc, and galactose, as UDP-Gal, are delivered into the Golgi apparatus by SLC35A3 and SLC35A2 transporters, respectively. However, although the UDP-Gal transporting activity of SLC35A2 has been clearly demonstrated, UDP-GlcNAc delivery by SLC35A3 is not fully understood. Therefore, we analyzed a panel of CHO, HEK293T, and HepG2 cell lines including WT cells, SLC35A2 knockouts, SLC35A3 knockouts, and double-knockout cells. Cells lacking SLC35A2 displayed significant changes in N- and O-glycan synthesis. However, in SLC35A3-knockout CHO cells, only limited changes were observed; GlcNAc was still incorporated into N-glycans, but complex type N-glycan branching was impaired, although UDP-GlcNAc transport into Golgi vesicles was not decreased. In SLC35A3-knockout HEK293T cells, UDP-GlcNAc transport was significantly decreased but not completely abolished. However, N-glycan branching was not impaired in these cells. In CHO and HEK293T cells, the effect of SLC35A3 deficiency on N-glycan branching was potentiated in the absence of SLC35A2. Moreover, in SLC35A3-knockout HEK293T and HepG2 cells, GlcNAc was still incorporated into O-glycans. However, in the case of HepG2 cells, no qualitative changes in N-glycans between WT and SLC35A3 knockout cells nor between SLC35A2 knockout and double-knockout cells were observed. These findings suggest that SLC35A3 may not be the primary UDP-GlcNAc transporter and/or different mechanisms of UDP-GlcNAc transport into the Golgi apparatus may exist.     

Constraints on the transport and glycosylation of recombinant IFN-gamma in Chinese hamster ovary and insect cells.

In this study we compare intracellular transport and processing of a recombinant glycoprotein in mammalian and insect cells. Detailed analysis of the N-glycosylation of recombinant human IFN-gamma by matrix-assisted laser-desorption mass spectrometry showed that the protein secreted by Chinese hamster ovary and baculovirus-infected insect Sf9 cells was associated with complex sialylated or truncated tri-mannosyl core glycans, respectively. However, the intracellular proteins were predominantly associated with high-mannose type oligosaccharides (Man-6 to Man-9) in both cases, indicating that endoplasmic reticulum to cis-Golgi transport is a predominant rate-limiting step in both expression systems. In CHO cells, although there was a minor intracellular subpopulation of sialylated IFN-gamma glycoforms identical to the secreted product (therefore associated with late-Golgi compartments or secretory vesicles), no other intermediates were evident. Therefore, anterograde transport processes in the Golgi stack do not limit secretion. In Sf9 insect cells, there was no direct evidence of post-ER glycan-processing events other than core fucosylation and de-mannosylation, both of which were glycosylation site-specific. To investigate the influence of nucleotide-sugar availability on cell-specific glycosylation, the cellular content of nucleotide-sugar substrates in both mammalian and insect cells was quantitatively determined by anion-exchange HPLC. In both host cell types, UDP-hexose and UDP-N-acetylhexosamine were in greater abundance relative to other substrates. However, unlike CHO cells, sialyltransferase activity and CMP-NeuAc substrate were not present in uninfected or baculovirus-infected Sf9 cells. Similar data were obtained for other insect cell hosts, Sf21 and Ea4. We conclude that although the limitations on intracellular transport and secretion of recombinant proteins in mammalian and insect cells are similar, N-glycan processing in Sf insect cells is limited, and that genetic modification of N-glycan processing in these insect cell lines will be constrained by substrate availability to terminal galactosylation.     

Complete enzymic synthesis of the mucin-type sialyl Lewis x epitope, involved in the interaction between PSGL-1 and P-selectin

Sialyl Lewis x (sLe(x)) is an established selectin ligand occurring on N- and O-linked glycans. Using a completely enzymic approach starting from p-nitrophenyl N-acetyl-alpha-D-galactosaminide (GalNAc(alpha1-pNp as core substrate, the sLe(x)-oligosaccharide Neu5Ac(alpha2-3)Gal(beta1-4)[Fuc(alpha1-3)]GlcNAc(beta1-6)[Gal(bet a1-3)]GalNAc(alpha1-pNp, representing the O-linked form, was synthesized in an overall yield of 32%. In a first step, Gal(beta1-3)GalNAc(alpha1-pNp was prepared in a yield of 52% using UDP-Gal and an enriched preparation of beta3-galactosyltransferase (EC 2.4.1.122) from rat liver. UDP-GlcNAc and a recombinant affinity-purified preparation of core 2 beta6-N-acetylglucosaminyltransferase (EC 2.4.1.102) fused to Protein A were used to branch the core 1 structure, affording GlcNAc(beta1-6)[Gal(beta1-3)]GalNAc(alpha1-pNp in a yield of >85%. The core 2 structure was galactosylated using UDP-Gal and purified human milk beta4-galactosyltransferase 1 (EC 2.4.1.38) (yield of >85%), then sialylated using CMP-Neu5Ac and purified recombinant alpha3-sialyltransferase 3 (EC 2.4.99.X) (yield of 87%), and finally fucosylated using GDP-Fuc and recombinant human alpha3-fucosyltransferase 6 (EC 2.4.1.152) produced in Pichia pastoris (yield of 100%). Overall 1.5 micromol of product was prepared. MALDI TOF mass spectra, and 1D and 2D TOCSY and ROESY 1H NMR analysis confirmed the obtained structure.