Abstract

Elevated ammonium concentrations in the medium of cultivated cells have been shown to increase the intracellular levels of uridine-5′-diphospho-N-acetylglucosamine (UDP-GlcNAc) and uridine-5′-diphospho-N-acetylgalactosamine (UDP-Gal-NAc; Ryll et al. , 1994 ). These sugar nucleotides are substrates for glycosyltransferases in the glycosylation pathway. In our experiments, recombinant Chinese hamster ovary cells producing an immunoadhesin glycoprotein (GP1-IgG) have been cultivated under controlled cell culture conditions in the presence of different ammonium concentrations. 15 N-Labeled ammonium chloride ( 15 NH 4 Cl) was added exogenously to the cell culture media to determine if ammonium was incorporated into UDP-GlcNAc and cytidine-5′-monophospho-N-acetylneuraminic acid (CMP-NANA) pools, and subsequently incorporated into GP1-IgG as N-linked glycans. The intracellular pools of UDP-activated hexosamines (UDP-GNAc) were followed during the time course of the experiment. To assess the extent of 15 NH 4 + incorporation into the glycans of GP1-IgG, the glycoprotein was first purified to homogeneity by protein A chromatography. Enzymatically released N-glycans were then analyzed by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. N-Glycans synthesized in the presence of 15 NH 4 Cl revealed an N-glycan-dependent increase in mass-to-charge of 2.5-4.8 Da. These results indicate that 60-70% of the total nitrogen containing monosaccharides had incorporated 15 N. Presumably, 15 NH 4 + was incorporated into GlcNAc and N-acetylneuraminic acid as proposed earlier ( Ryll et al. , 1994 ). This might be a universal and previously not described reaction in mammalian cells when exposed to nonphysiological but in cell culture commonly found concentrations of ammonium. The data presented here are of significance for glycoprotein production in mammalian cell culture, since it has been shown previously that elevated levels of UDP-activated hexosamines affect N-glycan characteristics such as branching and degree of amino sugar incorporation. In addition, our results demonstrate that isotope labeling in combination with MALDI-TOF-MS can be used as an alternate tool to radioactive labeling of sugar substrates in metabolic studies.

Introduction

Many recombinant proteins manufactured for use as pharmaceutical therapeutics are glycoproteins produced in mammalian cells. Oligosaccharide moieties are critical for numerous properties essential to fully functional glycoprotein therapeutics (e.g., in vivo clearance, solubility, immunogenicity, etc.; Varki, 1993 ). Protein glycosylation has been shown to be protein-and cell-specific ( Kornfeld and Kornfeld, 1985 ; Conradt et al. , 1990 ). Numerous studies have shown that environmental parameters also affect oligosaccharide structures on glycoproteins ( Goochee et al. , 1991 ; Chotigeat et al. , 1994 ; Gawlitzek et al. , 1995a , b ; Hooker et al. , 1995 ; Jenkins et al. , 1996 ). Product consistency is a prerequisite for product approval. To design efficient and reliable cell culture processes, the mechanisms by which culture conditions influence product quality need to be better understood.

Glutamine is a major energy source for permanent mammalian cells ( Zielke et al. , 1984 ). Ammonium is mainly a by-product of glutamine metabolism and thermal degradation of glutamine to D-pyroglutamic acid. In batch or fed batch cultures and depending upon the culture conditions (e.g., glutamine concentration), ammonium can accumulate to concentrations of 10 mM or more ( Moore et al. , 1997 ). Several of the effects of ammonium on cell growth, metabolism, nucleotide pools, and protein glycosylation have been reported previously (reviewed in Schneider et al. , 1996 ).

The concentration of ammonium in mammalian cell cultures has been shown to influence the composition of oligosaccharide structures observed. For example, sialylation of immunoglobulin in plasma cells was completely inhibited in the presence of 10 mM NH 4 Cl ( Thorens and Vassalli, 1986 ). CHO cells cultivated in the presence of elevated NH 4 Cl levels synthesized a larger amount of unglycosylated mouse placental lactogen-I than cells grown under low ammonium conditions ( Borys et al. , 1994 ). Ammonium concentrations ranging from 0-10 mM significantly reduced the sialylation of G-CSF O-glycans produced by recombinant CHO cells ( Andersen and Goochee, 1995 ). In recent studies we have shown that ammonium increased the antennarity of IL-2 N-glycosylation mutants synthesized in BHK-21 cells ( Gawlitzek et al. , 1998 ). Higher branching correlated with increased UDP-GNAc pools.

Previously we demonstrated a rapid dose-dependent increase in the intracellular UDP-N-acetylglucosamine and UDP-N-acetylgalactosamine levels in four different cell lines (BHK, CHO, Ltk-929, and hybridoma) in response to exogenously applied ammonium chloride ( Ryll et al. , 1994 ). These nucleotide sugars are substrates in the synthesis of N- and O-linked oligosaccharides on secreted or membrane bound glycoproteins ( Kornfeld and Kornfeld, 1985 ). Furthermore, UDP-GlcNAc is also a substrate for intracellular O-GlcNAcylation ( Haltiwanger et al. , 1992 ).

In principle ammonium can be incorporated into nucleotide sugars via two different pathways. A direct incorporation into fructose-6-P catalyzed by glucosamine-6-phosphate isomerase (GNPI, E.C. 5.3.1.10) or alternatively in a two-step mechanism by glutamine synthetase (GS, E.C. 6.3.1.2) and glucosamine-6-phosphate synthase (GNPS, E.C. 2.6.1.16.). GS activity has been shown to be regulated by the glutamine concentration in the medium of various mammalian cells, including CHO cells ( Miller et al. , 1978 ; Lacoste et al. , 1982 ; Feng et al. , 1990 ; Street et al. , 1993 ). GNPI activity has been studied in prokaryotes and eukaryotes (Benson and Friedman, 1970; Vogler et al. , 1989 ; Lara-Lemus et al. , 1992 ; Oliva et al. , 1995 ; Weidanz et al. , 1995 ; Wolosker et al. , 1998 ). Although GNPI is generally regarded as serving a catabolic function in the metabolism of glucosamine ( Weidanz et al. , 1995 ), the reaction catalyzed by this enzyme is reversible (Benson and Friedman, 1970; Vogler et al. , 1989 ; Lara-Lemus et al. , 1992 ). In previous studies we proposed the direct incorporation of ammonium into fructose-6-phosphate via GNPI to synthesize glucosamine-6-phosphate, a direct precursor of UDP-GlcNAc, and its significance for glycoprotein production in mammalian cell culture ( Ryll et al. , 1994 ).

Fig. 1.

Cultivation of recombinant CHO cells secreting GP1-IgG in the presence of different ammonium concentrations: 10 l reactor (diamonds), -Gln (open circles), -Gln/+ 15 NH 4 Cl (solid circles), +Gln (open squares), +Gln/+ 15 NH 4 Cl (solid squares). Culture conditions were changed on day 4 as described in detail under Materials and methods . Viable cell concentrations are average values from duplicate runs (n = 2). Error bars indicate 1 SD.

Fig. 1.

Cultivation of recombinant CHO cells secreting GP1-IgG in the presence of different ammonium concentrations: 10 l reactor (diamonds), -Gln (open circles), -Gln/+ 15 NH 4 Cl (solid circles), +Gln (open squares), +Gln/+ 15 NH 4 Cl (solid squares). Culture conditions were changed on day 4 as described in detail under Materials and methods . Viable cell concentrations are average values from duplicate runs (n = 2). Error bars indicate 1 SD.

In the present study we have demonstrated the incorporation of ammonium into the N-glycans of the recombinant immunoadhesin GP1-IgG ( Chamow and Ashkenazi, 1996 ) synthesized by CHO cells. Cells were cultivated in the presence or absence of 15 NH 4 Cl with and without glutamine. Intracellular nucleotide pools were measured during the process. In addition, enzymatically released N-glycans were analyzed by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF-MS).

Results

Production of GP1-IgG

CHO cells producing GP1-IgG were cultivated in the presence of different ammonium concentrations. Various ammonium levels were achieved by varying the glutamine and 15 NH 4 Cl levels. A primary 10 l bioreactor was used to inoculate eight 3 l bioreactors containing the respective media (as described under Materials and methods ; see Figure 1 ). No detrimental effect of high ammonium levels on cell growth and viability was observed. Cellular productivity of GP1-IgG was similar in all cases. The cell culture process was terminated on day 10 and secreted GP1-IgG purified by protein A chromatography.

Fig. 2.

Average ammonium concentration (n = 2, error bars indicate 1 SD) in different cell culture cases: 10 l reactor (diamonds), -Gln (open circles), -Gln/+ 15 NH 4 Cl (solid circles), +Gln (open squares), +Gln/+ 15 NH 4 Cl (solid squares). NH 4 + was measured by flow injection analysis.

Fig. 2.

Average ammonium concentration (n = 2, error bars indicate 1 SD) in different cell culture cases: 10 l reactor (diamonds), -Gln (open circles), -Gln/+ 15 NH 4 Cl (solid circles), +Gln (open squares), +Gln/+ 15 NH 4 Cl (solid squares). NH 4 + was measured by flow injection analysis.

Ammonium concentration

Ammonium levels were monitored throughout the experiment ( Figure 2 ). In the 10 l reactor, the ammonium concentration increased from a starting level of ∼0.9 mM on day 0 to about 6.5 mM on day 4. After medium exchange on day 4, the use of different media (plus/minus glutamine and plus/minus 15 NH 4 Cl) resulted in different ammonium levels. Glutamine-free medium resulted in the lowest ammonium concentrations (∼1.9 mM on day 10). About 5.5 mM ammonium was detected on day 10 in the glutamine-containing case. The ammonium concentration increased over time under both conditions. Significantly higher ammonium levels were reached in cultures where 15 NH 4 Cl was added (∼13 mM on day 4). In contrast, the NH 4 + concentration decreased slightly over time in the glutamine-free case (∼11 mM on day 10). A small increase was observed in glutamine-containing cultures (∼14 mM on day 10).

Intracellular UDP-GNAc pool

The intracellular UDP-GNAc pool was monitored by Ion-Pair-RP-HPLC ( Ryll and Wagner, 1991 ). Figure 3 shows the time course in response to different ammonium concentrations. Cell specific UDP-GNAc was stable during exponential growth and increased about 5-fold after the temperature shift to 31°C on day 2. The cell specific UDP-GNAc amount remained unchanged for 30 min after cells were suspended into the different media (first sample). Depending on the presence or absence of supplemented ammonium chloride, UDP-GNAc increased or decreased, respectively, during the following 20 h resulting in a 2-fold difference depending on ammonium concentration. After day 5, UDP-GNAc decreased in the high ammonium cultures with little difference between glutamine-containing and glutamine-free cases. With no added ammonium chloride in the background, the concentration of hexosamine sugars stayed low throughout day 9 in glutamine-free medium. However, glutamine containing medium resulted in an increase in hexosamine sugars after day 5 even without ammonium addition. After day 7, no difference could be found between the glutamine-containing/ammoniumfree and the ammonium supplemented cases.

Fig. 3.

Cell specific UDP-GNAc content (n = 4, error bars indicate 1 SD) of CHO cells cultivated with or without glutamine and 15 NH 4 Cl: 10 l reactor (diamonds), -Gln (open circles), -Gln/+ 15 NH 4 Cl (solid circles), +Gln (open squares), +Gln/+ 15 NH 4 Cl (solid squares). An intracellular content of 1 fmol per cell corresponds to an average intracellular concentration of 500 µM (assuming a cell volume of 2 · 10 -12 l).

Fig. 3.

Cell specific UDP-GNAc content (n = 4, error bars indicate 1 SD) of CHO cells cultivated with or without glutamine and 15 NH 4 Cl: 10 l reactor (diamonds), -Gln (open circles), -Gln/+ 15 NH 4 Cl (solid circles), +Gln (open squares), +Gln/+ 15 NH 4 Cl (solid squares). An intracellular content of 1 fmol per cell corresponds to an average intracellular concentration of 500 µM (assuming a cell volume of 2 · 10 -12 l).

N-Glycan analysis by MALDI-TOF-MS

To determine if 15 NH 4 + was incorporated into the N-glycans of the glycoprotein product, and if supplementing the culture media with ammonium chloride increased the branching of the N-linked glycans, enzymatically released N-glycans were analyzed by MALDI-TOF-MS ( Papac et al. , 1996 ). GP1-IgG produced with cell culture conditions containing normal levels of ammonium displayed N-linked glycans with 60% of the N-linked oligosaccharides as biantennary, 30% triantennary, and 10% tetraantennary structures (data not shown). No change in branching of the N-glycans was observed in response to increased ammonium concentrations.

All N-glycans synthesized in cultures supplemented with 15 NH 4 Cl revealed a significant positive mass shift when compared to the same structures produced in control cultures ( Figure 4 ). The native N-linked glycans were analyzed in the negative-ion mode to determine total incorporation of 15 NH 4 , desialylated oligosaccharides were analyzed in the positive-ion mode to differentiate incorporation into N-acetylglucosamine and N-acetylneuraminic acid. In desialylated glycans, the increase in m/z correlated with the content of N-acetylglucosamine, but was unaffected by the content of galactose as would be expected ( Table I and II ). Sialylated oligosaccharides showed a larger mass shift than their desialylated counterparts ( Table I and II ). These observations indicate that both N-acetylglucosamine (GlcNAc) and N-acetylneuraminic acid were labeled with 15 N. The MALDI-TOF data suggest that a greater percentage of 15 N was incorporated into N-acetylglucosamine than into N-acetylneuraminic acid ( Table I and II , p = 0.002). No significant difference in the average mass shift between the glutamine-free + 15 NH 4 Cl culture and the glutamine-containing + 15 NH 4 Cl case was observed (p = 0.59). Approximately 60-70% of the total nitrogen-containing monosaccharides were labeled with 15 N. The precision of the mass measurements decreases as the signal-to-noise of the peak being measured decreases. The worst precision was obtained for the least abundant glycan species (e.g., 3100, 3121, 3131, and 3132; Table II ). Even in the worst cases, the error in the calculated amount of 15 N incorporation was not greater than 11%.

Fig. 4.

MALDI-TOF mass spectra of PNGase F released N-linked oligosaccharides from GP1-IgG showing examples for mass differences of N-glycans synthesized in the absence (solid line) or presence (dotted line) of 15 NH 4 Cl. ( A ) Glutamine free medium: neutral N-glycans after desialylation measured in positive ion mode. ( B ) Glutamine containing medium: sialylated structures measured in negative ion mode. N-Glycan peaks are assigned by a four-digit number code (*; described in Table I and II footnote) and by the measured mass. Mass differences of compared structures are given in brackets.

Fig. 4.

MALDI-TOF mass spectra of PNGase F released N-linked oligosaccharides from GP1-IgG showing examples for mass differences of N-glycans synthesized in the absence (solid line) or presence (dotted line) of 15 NH 4 Cl. ( A ) Glutamine free medium: neutral N-glycans after desialylation measured in positive ion mode. ( B ) Glutamine containing medium: sialylated structures measured in negative ion mode. N-Glycan peaks are assigned by a four-digit number code (*; described in Table I and II footnote) and by the measured mass. Mass differences of compared structures are given in brackets.

Discussion

An increase in the intracellular concentration of UDP-activated hexosamine sugars (UDP-GNAc) has been shown to be the result of factors that affect cell growth rate including nutrient exhaustion ( Ryll and Wagner 1992 ), temperature shift ( Moore et al. , 1997 ), and the addition of glucosamine or ammonium chloride ( Ryll et al. , 1994 ; Gawlitzek et al. , 1998 ). UDP-activated hexosamines are precursors for N- and O-linked glycosylation as well as O-GlcNAcylation. In recent studies it has been shown that elevated UDP-GNAc levels lead to increased branching of N-linked oligosaccharide structures ( Gawlitzek et al. , 1998 ) suggesting an increased activity of UDP-GlcNAc transferases IV and/or V. Furthermore, Schlenke et al. (1997) reported that BHK-21A cells synthesize N-glycans containing a mixture of the typical N-acetyllactosamine motif (Gal(β1-4)GlcNAc-R) together with unsialylated GalNAc(β1-4)GlcNAc-R moieties. In BHK-21A cells producing recombinant erythropoietin, the addition of ammonium chloride resulted in an increased incorporation of GalNAc instead of galactose as a result of an enlarged UDP-GalNAc pool (H.S.Conradt, personal communication). Due to the incapability of the endogenous α2,3-sialyltransferase to transfer NANA to the GalNAc(β1-4)GlcNAc-R structure, here the ammonium level also influenced the degree of sialylation.

In our experiment, the concentrations of UDP-GNAc also increased as a response to the growth arrest by temperature shift (from 37°C to 31°C). High UDP-GNAc bearing cells were transferred into fresh media plus or minus glutamine and with or without ammonium chloride. Cells containing high concentrations of UDP-GNAc deplete that pool quickly after being transferred into fresh growth medium at 37°C (data not shown). However, at 31°C, cells are in a non- or slow-growing condition as has been shown previously for similar cultures shifted to 30°C ( Moore et al. , 1997 ). Under these conditions the UDP-GNAc pool was not reduced to levels typical for this cell line at exponential growth even with very little ammonium in background (Gln-free case). The accumulation of ∼6 mM NH 4 + in the glutamine-containing culture resulted in an increase of UDP-GNAc after day 5. In contrast, the intracellular concentration of UDP-GNAc increased immediately after cells were transferred into medium containing ammonium chloride.

The results presented in this article indicate that 15 N from supplemented 15 NH 4 Cl has been incorporated into N-glycans of GP1-IgG via UDP-GlcNAc and CMP-NANA. No significant difference in the incorporation of 15 N was found between the glutamine-containing and glutamine-free cases ( Table I and II ). Since GS activity has been shown to be regulated by the glutamine concentration in the medium, our findings suggest that under the conditions used here ammonium is mainly incorporated by GNPI into Frc-6-P to form GlcN-6-P. Subsequently, GlcN-6-P becomes acetylated and finally activated by UTP resulting in UDP-GNAc. UDP-GlcNAc is a direct precursor in the formation of CMP-NANA ( Warren, 1972 ; Pels Rijcken et al. , 1995 ). Consequently, if GlcNAc is labeled with 15 N due to the incorporation of 15 NH 4 +, NANA also has to be labeled with 15 N. This assumption was confirmed by our data, since sialylated N-glycans revealed a larger mass shift compared to their unsialylated equivalents. To our knowledge this finding is the first reported evidence that GNPI activity can play an important role in the biosynthesis of recombinant glycoproteins in mammalian cell cultures.

In contrast to results observed in previous studies using IL-2 N-glycosylation variants as model glycoproteins ( Gawlitzek et al. , 1995b ), no effect of ammonium on branching of N-glycans was observed here (data not shown). Cell- and protein-specific glycosylation as well as different culture conditions could be an explanation for the different results. Different cell lines (BHK-21 vs. CHO) were used in the two experiments. The transport of activated nucleotide sugars might be regulated differently in these cells. Another important fact is that the accessibility of N-glycosylation sites is determined by the 3-D structure of the protein. Thus, one protein might be affected more easily (e.g., IL-2 N-glycosylation mutants) by culture conditions than others (e.g., GP1-IgG). The different mode of cultivation (perfusion culture vs. batch) in the two experiments, which had significant effects on the UDP-GNAc pool, could be another explanation. The difference between glutamine-free cultures and those supplemented with ammonium chloride were much more pronounced in the perfusion culture (6-fold increase vs. 2-fold reported here). However, the intracellular UDP-GNAc amounts were significantly higher in the CHO cells used in this study (5-10 vs. 0.25-1.5 fmol/cell for CHO and BHK cells, respectively).

O-GlcNAcylation of Ser (Thr) sites in many cellular proteins has been shown to be involved in a variety of processes such as transcriptional initiation (for a review, see Hart et al. , 1996 ). Since ammonium is known to influence cell growth and other cellular functions and it promotes the accumulation of UDP-GlcNAc, it cannot be ruled out that accumulation of ammonium in cell cultures could influence such O-GlcNAylation reactions. This could be an additional mechanism by which ammonium exhibits its effect on cultivated mammalian cells beside widely discussed mechanisms of ammonium action (for review, see Schneider et al. , 1996 ). In that context it is interesting to note that O-GlcNAc transferase might have a biphasic Km for UDP-GlcNAc (R.S.Haltiwanger, personal communication) leading to a second activity increase with an apparent Km around 2.5 mM. Growing mammalian cells usually contain UDP-GlcNAc in the high micromolar range. A high ammonium concentration can lead to an increase to 2.5 mM or more as shown here (see Figure 3 ; a typical cell volume for CHO cells is 2-3 × 10 -12 l) and previously for BHK, CHO, L929, and hybridoma cells ( Ryll et al. 1994 ). Experiments addressing O-GlcNAcylation should therefore take into account that the intracellular precursor pool can vary drastically depending on growth state and media condition of the particular cell culture being used.

Our results verify that 15 N labeled ammonium present in cell culture medium can be detected in N-linked oligosaccharides of recombinant glycoproteins and therefore must be incorporated into nucleotide sugar pools. These findings demonstrate that 15 N labeling together with mass spectrometry analysis of carbohydrates is a useful technique to investigate the biochemical pathway of oligosaccharide synthesis in recombinant protein glycosylation. It is also a powerful tool to investigate the pleiotropic effects of ammonium in animal cell culture.

Materials and methods

Cells

Experiments were performed using Chinese hamster ovary (CHO) cells derived from a dihydrofolate reductase minus (dhfr - ) CHO DUKX B11 host ( Urlaub and Chasin, 1980 ). Cells were genetically engineered to secrete GP1-IgG using a dhfr/methotrexate selection method similar to that used by Kaufman and Sharp (1982) .

Bioreactor experiments

All experiments were performed in stirred tank bioreactors (Applikon, Foster City, CA) with marine vortexing impellers. Reactors were equipped with calibrated dissolved oxygen, pH and temperature probes. Dissolved oxygen was controlled on-line through sparging with air and/or oxygen at 60 ± 5% air saturation. pH was maintained through additions of CO 2 or Na 2 CO 3 at pH 7.2 ± 0.1. Vessels were equipped with temperature jackets and cells were grown at 37°C for the first 48 h. After 2 days the cultivation temperature was reduced to 31°C. Cell viability was assessed by trypan blue exclusion. Cells were inoculated at a concentration of 1.2 × 10 6 ml -1 . To ensure equal starting conditions for different cases, the experiment was started in a 10 l reactor with 8.5 l working volume. On day 4 cells were harvested by centrifugation at 150 × g for 10 min. Cells were resuspended in the following media: (1) glutamine-free + 13 mM NaCl, (2) glutamine-free + 13 mM 15 NH 4 Cl (Sigma), (3) glutamine containing + 13 mM NaCl, (4) glutamine containing + 13 mM 15 NH 4 Cl (all cases in duplicates). Eight 3 l bioreactors with 1.5 l working volume were inoculated at a concentration of 5 × 10 6 ml -1 . Cultures were terminated on day 10.

Ammonium measurement

The ammonium concentration in daily cell culture samples was measured by flow-injection analysis similar to a system described earlier ( Forman et al. , 1991 ).

Nucleotide sugar extraction and analysis

Intracellular nucleotide pools were analyzed as described previously ( Ryll et al. , 1991 ). Briefly, 2-6 × 10 6 cells were cooled and pelleted at 200 × g for 3 min. Cell pellets were extracted using 0.5 M perchloric acid. Supernatants were pooled after repeating the extraction step, and neutralized using 2.5 M KOH in 1.5 M K 2 HPO 4 . The clear filtered supernatant was used for quantification using an ion-pair reversed phase HPLC method running on an HP1090 system (Hewlett Packard, Mountain View, CA). The following buffers (A: 0.1 M KH 2 PO 4 /K 2 HPO 4 + 8 mM tetrabutylammonium hydrogensulfate pH 5.3; B: 70 (v/v) % buffer A + 30 (v/v) % methanol, pH 5.9), elution gradient and flow rates were used: 0-2.5 min 0% B at 1.3 ml/min, 2.5-16 min 0-40% B at 1.3 ml/min, 16-18 min 40-55% B at 1.3 ml/min, 18-19.5 min 55-65% B at 1.0 ml/min, 19.5-20 min 65-100% B at 1.0 ml/min, 20-26 min 100% B at 1.0 ml/min, 26-26.5 min 100-0% B at 1.0 ml/min, 26.5-27.5 min 0% B at 1.0 ml/min, 27.5-36 min 0% B at 1.3 ml/min. Quantification was done by integrating peak areas using calibration curves for all nucleotides in the appropriate range.

Purification of GP1-IgG

Cell suspension was harvested and cells removed by low speed centrifugation (200 g for 10 min). Cell free supernatant from each case was microfiltered (0.2 µm). GP1-IgG was purified by protein A chromatography on FPLC (Pharmacia LKB). Flow rates were 4 ml/min for equilibration, load and wash steps, 2 ml/min for elution. A ProSep Protein A column (1 ml bed volume, Bioprocessing Ltd., Durham, England) was equilibrated with PBS (pH 7.0) and an appropriate volume of supernatant was loaded. The column was washed with PBS until baseline was reached and GP1-IgG was eluted with sodium citrate (60 mM, pH 3.0). Afterwards, protein A was regenerated with a buffer containing 20 mM Tris and 2 M guanidine HCl (pH 7.5).

N-Glycan release and desialylation

The protein A purified GP1-IgG, ranging in concentration from 1.9 to 2.6 mg/ml, was buffer exchanged into reduction and carboxymethylation buffer containing 8 M urea, 360 mM Tris, and 3.2 mM EDTA (pH 8.6) using a Centricon 30 (Amicon Inc.) ultrafiltration device. Reduction was performed with 10 mM dithiothreitol for 4 h at 37°C. The reduced protein was then alkylated with 25 mM iodoacetic acid by incubation in the dark at 37°C for 30 min. The reduced and carboxymethylated GP1-IgG was dialyzed into 100 mM ammonium bicarbonate and then lyophilized to dryness. The lyophilized protein was dissolved in PNGase F digestion buffer (75 mM sodium phosphate (pH 8.0) containing 5 mM EDTA and 0.02% sodium azide) to a final concentration of 2.5 mg/ml. The N-linked oligosaccharides were released by incubation with (6.4 units/mg protein) glycerol-free PNGase F (Boehringer Mannheim) overnight at 37°C. The released oligosaccharides were recovered in the supernatant following protein precipitation with 75% ice-cold ethanol. The recovered oligosaccharides were hydrolyzed to the free reducing oligosaccharides by treatment with 13 mM acetic acid at room temperature for 2 h. Following hydrolysis, the oligosaccharides were dried in a Savant Speed Vac, and were reconstituted with deionized water to a final concentration equivalent to 2 mg/ml (based upon the original protein concentration). A portion of the released glycans were desialyalted by treatment with 10% glacial acetic acid at 80°C for 3 h. Following desialylation sialic acid was removed with 0.25 ml of AG1-X2 resin acetate form (Bio-Rad).

Sample preparation for MALDI-TOF

Prior to MALDI-TOF analysis, all samples were desalted with 0.25 ml of AG50W-X8 resin (hydrogen form, Bio-RAD) to remove cations. The 2,5-dihydroxybenzoic acid matrix (sDHB) was prepared by dissolving 5 mg of 2,5-dihydroxybenzoic acid + 0.25 mg of 5-methoxysalicylic acid in 1 ml of ethanol/10 mM aqueous sodium chloride 1:1 (v/v). The 2′,4′,6′,-trihydroxyacetophenone matrix (THAP) was prepared by dissolving 1 mg of 2′,4′,6′,-trihydroxyacetophenone in 1 ml of acetonitrile/20 mM ammonium citrate 1:1 (v/v). For the desialylated oligosaccharides, 0.5 µl of analyte was mixed with 0.5 µl of sDHB. For the sialylated oligosaccharides, 0.5 µl of analyte was mixed with 0.75 µl of THAP. All samples were dried under vacuum (50 × 10 -3 Torr).

Instrument operating parameters

The MALDI-TOF mass spectrometer used to acquire the spectra was a Voyager Elite (PerSeptive Biosystems, Framingham, MA). All samples were irradiated with UV light (337 nm) from a N 2 laser. Desialylated oligosaccharides were analyzed at 25 kV with a single-stage reflectron in the positive-ion mode. Sialylated oligosaccharides were analyzed at 20 kV without the reflectron, but with delayed extraction (100 ns) in the negative-ion mode.

A two-point external calibration was used for mass assignment of the desialylated oligosaccharides. The calibrants used for the positive-ion mode were a high mannose oligosaccharide [(M+Na) +avg = 1258.1 m/z ] and a triantennary complex oligosaccharide [(M+Na) +avg = 2029.83 m/z ]. When the sialylated oligosaccharides were analyzed, a two-point internal calibration containing disialyl tetraose [(M-H) -avg = 964.86 m/z ] and disialylated biantennary [(M-H) -avg = 2019.82 m/z ] was used to calibrate the instrument. Generally, a mass accuracy of <0.005% was achieved in the reflector mode of operation using external calibration, and a mass accuracy of <0.006% was obtained in the linear mode using internal calibration.

Typically, spectra from 64 to 128 laser shots were summed to obtain the final spectrum. All the spectra were smoothed 19 points with a Savitsky-Golay function. To prevent matrix ions from saturating the multichannel plate detector, the detector voltage was held below threshold until ions of mass-to-charge greater than 500 m/z could strike the detector.

Acknowledgments

We gratefully acknowledge the excellent technical assistance of Denise Woon, Blage Zelalich, and Derf Lewis.

Abbreviations

    Abbreviations
  • BHK

    baby hamster kidney

  • CHO

    Chinese hamster ovary

  • FPLC

    fast protein liquid chromatography

  • MALDI-TOF-MS

    matrix-assisted laser desorption ionization/time-of-flight mass spectrometry

  • Frc-6-P

    fructose-6-phosphate

  • GlcN-6-P

    glucosamine-6-phosphate

  • GNPI

    glucosamine-6-phosphate isomerase

  • GS

    glutamine synthetase

  • GNPS

    glucosamine-6-phosphate synthase

  • NANA

    N-acetylneuraminic acid

  • PBS

    phosphate-buffered saline

  • UDP-GalNAc

    UDP-N-acetylgalactosamine

  • UDP-GlcNAc

    UDP-N-acetylglucosamine

  • UDP-GNAc

    sum of UDP-GalNAc and UDP-GlcNAc

References

Andersen
D.C.
Goochee
C.F.
The effect of ammonia on the O-linked glycosylation of granulocyte colony-stimulating factor produced by chinese hamster ovary cells
Biotech. Bioeng.
 , 
1995
, vol. 
47
 (pg. 
96
-
105
)
Benson
R.L.
Friedman
S.
Allosteric control of glucosamine phosphate isomerase from adult housefly and its role in the synthesis of glucosamine 6-phosphate
J. Biol. Chem.
 , 
1969
, vol. 
245
 (pg. 
2219
-
2228
)
Borys
M.C.
Linzer
D.I.H.
Papoutsakis
T.
Ammonia affects the glycosylation patterns of recombinant mouse placental lactogen-I by chinese hamster ovary cells in pH-dependent manner
Biotechnol. Bioeng.
 , 
1994
, vol. 
43
 (pg. 
505
-
514
)
Chamow
S.M.
Ashkenazi
A.
Immunoadhesins: principles and applications
Trends Biotechnol.
 , 
1996
, vol. 
14
 (pg. 
52
-
60
)
Chotigeat
W.
Watanapokasin
Y.
Mahler
S.
Gray
P.P.
Role of environmental conditions on the expression levels, glycoform pattern and levels of sialyltransferase for hFSH produced by recombinant CHO cells
Cytotechnology
 , 
1994
, vol. 
15
 (pg. 
217
-
221
)
Conradt
H.S.
Department of Gene Regulation and Differentiation
Braunschweig, Germany
GBF
 
personal communication
Conradt
H.S.
Hofer
B.
Hauser
H.
Expression of human glycoproteins in recombinant mammalian cells: towards genetic engineering of N- and O-glycoproteins
Trends Glycosci. Glycotech.
 , 
1990
, vol. 
2
 (pg. 
168
-
180
)
Feng
B.
Shiber
S.K.
Max
S.R.
Glutamine reulates glutamine synthetase expression in skeletal muscle cells in culture
J. Cell. Physiol.
 , 
1990
, vol. 
98
 (pg. 
155
-
162
)
Forman
L.W.
Thomas
B.D.
Jacobson
F.S.
On-line monitoring and control of fermentation processes by flow-injection analysis
Analytica Chimica Acta
 , 
1991
, vol. 
249
 (pg. 
101
-
111
)
Gawlitzek
M.
Valley
U.
Nimtz
M.
Wagner
R.
Conradt
H.S.
Characterization of changes in the glycosylation pattern of recombinant proteins from BHK-21 cells due to different culture conditions
J. Biotechnol.
 , 
1995
, vol. 
42
 (pg. 
117
-
131
)
Gawlitzek
M.
Valley
U.
Nimtz
M.
Wagner
R.
Conradt
H.S.
Beuvery
E.C.
Griffiths
J.B.
Zeijlemaker
W.P.
Effects of ammonia and glucosamine on the glycosylation pattern of recombinant proteins expressed from BHK-21 cells
Animal Cell Technology: Developments Towards the 21st Century
 , 
1995
The Netherlands
Kluwer Academic
(pg. 
379
-
384
)
Gawlitzek
M.
Valley
U.
Wagner
R.
Ammonium ion and glucosamine dependent increase of oligosaccharide complexity in recombinant glycoproteins secreted from cultivated BHK-21 cells
Biotechnol. Bioeng.
 , 
1998
, vol. 
57
 (pg. 
518
-
528
)
Goochee
C.F.
Gramer
M.J.
Andersen
D.C.
Bahr
J.B.
Rasmussen
J.R.
The oligosaccharides of glycoproteins: bioprocess factors affecting oligosaccharide structure and their effect on glycoprotein properties
Bio/Technology
 , 
1991
, vol. 
9
 (pg. 
1347
-
1355
)
Haltiwanger
R.S.
Department of Biochemistry and Cell Biology
Stony Brook
State University of New York
 
personal communication, preliminary results
Haltiwanger
R.S.
Blomberg
M.A.
Hart
G.W.
Glycosylation of nuclear and cytoplasmic proteins
J. Biol. Chem.
 , 
1992
, vol. 
13
 (pg. 
9005
-
9013
)
Hart
G.W.
Kreppel
L.K.
Comer
F.I.
Arnold
C.S.
Snow
D.M.
Ye
Z.
Cheng
X.
DellaManna
D.
Caine
D.S.
Earles
B.J.
Akimoto
Y.
Cole
R.N.
Hayes
B.K.
O-GlcNAcylation of key nuclear and cytoskeletal proteins: reciprocity with O-phosphorylation and putative roles in protein multimerization
Glycobiology
 , 
1996
, vol. 
6
 (pg. 
711
-
716
)
Hooker
A.D.
Goldman
M.H.
Markham
N.H.
James
D.C.
Ison
A.P.
Bull
A.T.
Strange
P.G.
Salmon
I.
Baines
A.J.
Jenkins
N.
N-glycans of recombinant human interferon-γ change during batch cultures of chinese hamster ovary cells
Biotechnol. Bioeng.
 , 
1995
, vol. 
48
 (pg. 
639
-
648
)
Jenkins
N.
Parekh
R.B.
James
D.C.
Getting the glycosylation right: implications for the biotechnology industry
Nature Biotechnol.
 , 
1996
, vol. 
14
 (pg. 
975
-
981
)
Kaufman
R.J.
Sharp
P.A.
Amplification and expression of sequences cotransfected with a modular dihydrofolate reductase complementary DNA gene
J. Mol. Biol.
 , 
1982
, vol. 
159
 (pg. 
601
-
621
)
Kornfeld
R.
Kornfeld
S.
Assembly of asparagine-linked oligosaccharides
Annu. Rev. Biochem.
 , 
1985
, vol. 
54
 (pg. 
631
-
664
)
Lacoste
L.
Chaudhary
K.D.
Lapointe
J.
Derepression of the glutamine synthetase in neuroblastoma cells at low concentrations of glutamine
J. Neurochem.
 , 
1982
, vol. 
39
 (pg. 
78
-
85
)
Lara-Lemus
R.
Libreros-Minotta
C.A.
Altamirano
M.M.
Calgano
M.L.
Purification and characterization of glucosamine-6-phosphate deaminase from dog kidney cortex
Arch. Biochem. Biophys.
 , 
1992
, vol. 
297
 (pg. 
213
-
220
)
Miller
R.E.
Hackenberg
R.
Gershman
H.
Regulation of glutamine synthetase in cultured T3T-L1 cells by insuline, hydrocortisone and dibutyryl cyclic AMP
Proc. Natl. Acad. Sci. USA
 , 
1978
, vol. 
75
 (pg. 
1418
-
1422
)
Moore
A.
Mercer
J.
Dutina
G.
Donahue
C.J.
Bauer
K.D.
Mather
J.P.
Etcheverry
T.
Ryll
T.
Effects of temperature shift on cell cycle, apoptosis and nucleotide pools in CHO batch cultures
Cytotechnology
 , 
1997
, vol. 
23
 (pg. 
47
-
54
)
Oliva
G.
Fontes
M.R.M.
Garratt
R.C.
Altamirano
M.M.
Calcagno
M.L.
Horjales
E.
Structure and catalytic mechanism of glucosamine 6-phosphate deaminase from Escherichia coli at 2.1 Å resolution
Structure
 , 
1995
, vol. 
3
 (pg. 
1323
-
1332
)
Papac
D.I.
Wong
A.
Jones.
A.J.S.
Analysis of acidic oligosaccharides and glycopeptides by matrix-assisted laser desorption/ionization time-offlight mass spectrometry
Anal. Chem.
 , 
1996
, vol. 
68
 (pg. 
3215
-
3223
)
Pels Rijcken
W.R.
Overdijk
B.
Van Den Eijnden
D.H.
Ferwerda
W.
The effect of increasing nucleotide-sugar concentrations on the incorporation of sugars into glycoconjugates in rat hepatocytes
Biochem. J.
 , 
1995
, vol. 
305
 (pg. 
865
-
870
)
Ryll
T
Wagner
R.
Improved ion pair high-performance liquid chromatographic method for the quantification of a wide variety of nucleotides and sugar-nucleotides in animal cells
J. Chromatogr.
 , 
1991
, vol. 
570
 (pg. 
77
-
88
)
Ryll
T.
Wagner
R.
Intracellular ribonucleotide pools as a tool for monitoring the physiological state of in vitro cultivated mammalian cells during production process
Biotechnol. Bioeng.
 , 
1992
, vol. 
40
 (pg. 
934
-
946
)
Ryll
T.
Valley
U.
Wagner
R.
Biochemistry of growth inhibition by ammonium ions in mammalian cells
Biotechnol. Bioeng.
 , 
1994
, vol. 
44
 (pg. 
184
-
193
)
Schlenke
P.
Grabenhorst
E.
Wagner
R.
Nimtz
M.
Conradt
H.S.
Carrondo
M.J.T.
, et al.  . 
Expression of human a2,6-sialyltransferase in BHK-21 A cells increases the sialylation of coexpressed human erythropoietin: NeuAc-transfer onto GalNAc(β1-4)GlcNAc-R motifs
Animal Cell technology: From Vaccines to Genetic Medicine
 , 
1997
The Netherlands
Kluwer Academic
(pg. 
475
-
480
)
Schneider
M.
Marison
I.W.
von Stockar
U.
The importance of ammonia in mammalian cell cultures
J. Biotechnol.
 , 
1996
, vol. 
46
 (pg. 
161
-
185
)
Street
J.C.
Delort
A.-M.
Braddock
P.S.
Brindle
K.M.
A 1 H/ 15 N NMR study of nitrogen metabolism in cultured mammalian cells
Biochem. J.
 , 
1993
, vol. 
291
 (pg. 
485
-
492
)
Thorens
B.
Vassalli
P.
Chloroquine and ammonium chloride prevent terminal glycosylation of immunoglobulins in plasma cells without affecting secretion
Nature
 , 
1986
, vol. 
321
 (pg. 
618
-
620
)
Urlaub
G.
Chasin
L.A.
Isolation of Chinese hamster cell mutants deficient in dihydrofolate reductase activity
Proc. Natl. Acad. Sci. USA
 , 
1980
, vol. 
77
 (pg. 
4216
-
4220
)
Varki
A.
Biological roles of oligosaccharides: all of the theories are correct
Glycobiology
 , 
1993
, vol. 
3
 (pg. 
97
-
130
)
Vogler
A.P.
Trentmann
S.
Lengeler
J.W.
Alternative route for biosynthesis of amino sugars in Escherichia coli K-12 mutants of a catabolic isomerase
J. Bacteriol.
 , 
1989
, vol. 
171
 (pg. 
6586
-
6592
)
Warren
L.
Gottschalk
A.
The biosynthesis and metabolism of amino sugars and amino sugar-containing heterosaccharides
Glycoproteins
 , 
1972
Amsterdam
Elsevier
(pg. 
1097
-
1126
)
Weidanz
J.A.
Campbell
P.
DeLucas
L.J.
Jin
J.
Moore
D.
Rodén
L.
Yu
H.
Heilmann
E.
Vezza
A.C.
Glucosamine 6-phosphate deaminase in normal human erythrocytes
Br. J. Haematol.
 , 
1995
, vol. 
91
 (pg. 
72
-
79
)
Wolosker
H.
Kline
D.
Bian
Y.
Blackshaw
S.
Cameron
A.M.
Fralich
T.J.
Schnaar
R.L.
Snyder
S.H.
Molecularly cloned mammalian glucosamine-6-phosphate deaminase localizes to transporting epithelium and lacks oscillin activity
FASEB J.
 , 
1998
, vol. 
12
 (pg. 
91
-
99
)
Zielke
H.R.
Zielke
C.L.
Ozand
P.T.
Glutamine: a major energy source for cultured mammalian cells
Fed. Proc.
 , 
1984
, vol. 
43
 (pg. 
12
-
125
)

Author notes

2
Cell Culture & Fermentation R&D
3
Analytical Chemistry