Blotechnol. hog. 1994, 10, 39-44
39
Comparison of N-Linked Oligosaccharides of Recombinant Human Tissue Kallikrein Produced by Chinese Hamster Ovary Cells on Microcarrier Beads and in Serum-Free Suspension Culture Eric Watson,’ Bhavana Shah, Lisa Leiderman, Yueh-Rong HSU,Subhash Karkare, Hsieng S. Lu, and Fu-Kuen Lin &en
Inc., Amgen Center, Thousand Oaks, California 91320
Glycosylationheterogeneity in recombinant human tissue kallikrein (r-HuTK) produced by Chinese hamster ovary (CHO) cells from microcarrier culture and from a serum-free suspension cell recycle process has been compared. Significant differences in the degree of sialylation were observed in glycoform distribution and oligosaccharide heterogeneity. High-performance liquid chromatography with a pellicular anion-exchangecolumn under low pH eluant conditions was used to characterize the number and types of N-linked complex type oligosaccharides present. The oligosaccharides were released by N-glycanase and, after reduction, were resolved into a number of peaks containing one, two, three, and four sialic acids with an additional subfractionation based on the nature of the antennary structure. The microcarrier process resulted in a reduced amount of sialylated oligosaccharide species as compared to the suspension cell process. Removal of sialic acid followed by chromatography of the asialooligosaccharides under high pH anion-exchange conditions indicated that the same antennary structures were present but in slightly different relative amounts. The oligosaccharide profiles are indicative of a highly complex array of microheterogeneity present, encompassing mono-, di-, tri-, and tetrasialylated complex type oligosaccharides.
Introduction Tissue kallikreins are a closely related family of serine proteinases that release a potent peptide, Lys-bradykinin, by proteolysis from high or low molecular weight kininogens (Erdos and Wilde, 1979; Irie et al., 1982; Pisano and Austen, 1976). Tissue kallikreins occur mainly in the pancreas, pancreatic juice, salivary glands, saliva, kidney, and urine. These proteinases are acidic glycoproteins,and while the polypeptide chains in kallikreins isolated from different tissues of a given species have been shown to be closely related, they differ from each other in their carbohydrate content (Fiedler et al., 1981). A series of reports has described the amino acid sequence of human urinary kallikrein (Kellerman et al., 1988; Takahashi et al., 1988; Lu et al., 1989), but as yet there has been no information reported on the nature and type of oligosaccharides present. Two of these reports identified the sequence, Asn-X-Thr(Ser),at positions 78-80,84-86, and 141-143 (Takahashi et al., 1988; Lu et al., 19891, which is a common site for N-glycosylation, but did not find any evidence for 0-glycosylation. In contrast to these results, the same N-glycosylation sites were identified, but in addition, three 0-glycosylation sites were found linked to serine-threonine at positions 69,80, and 143 (Kellerman et al., 1989). There is no information regarding the identity and nature of oligosaccharides on human kallikrein. Recently, the human tissue kallikrein gene has been cloned and expressed in Chinese hamster ovary cells (Lin and Lu, 1989) in amounts sufficient for an evaluation of the type of glycosylationpresent on recombinant human tissue kallikrein (r-HuTK). In this report on r-HuTK, preliminary information is presented showing that the major oligosaccharides are
* Author to whom correspondence should be addressed. 875&7938/94/3010-0039$04.50/0
sialylated N-linked complex type structures. There was no evidence for significant amounts of 0-linked or high mannose type oligosaccharides. The various contributions of microcarrier culture and a serum-free suspension cell recycle process are compared as sources of variability in affecting r-HuTK glycosylation heterogeneity. Experimental Procedures Materials. Recombinant human tissue kallikreins are products of Amgen (Thousand Oaks, CAI. Fifty percent (w/w) sodium hydroxide solution was purchased from Fisher Scientific (Rockville, MD). Sodium acetate was reagent grade obtained from J. T. Baker (Philipsburg,NJ). N-Glycanase was obtained from Genzyme (Boston, MA). p-Aminobenzoic acid ethyl ester (ABEE) was purchased from Sigma Chemical Co. (St.Louis, MO). Neuraminidase (Arthrobacterureafaciens),specific activity ca. 100units/ mg of enzyme protein, was obtained from BoehringerMannheim Biochemicals (Indianapolis, IN). Media Composition. (1) Growth medium (for the microcarrier process): DMEM/F12 (Gibco, 875118EH) supplemented with 2 mM glutamine, 220 pg/mL serine, 65 pg/mL aspartate, 80 pg/mL asparagine, 250 pg/mL cysteine, 0.05 mM 2-mercaptoethanol, and 5% FBS. (2) Production medium 1 (for the microcarrier process): DMEM/F12 supplemented with 2 mM glutamine, 220 pg/mL serine, 65 pg/mL aspartate, 80 pg/mL asparagine, 250 pg/mL cysteine, 0.05 mM 2-mercaptoethanol, 5 pg/mL insulin, 5 pg/mL transferrin, 5 ng/mL sodium selenite, 20 ng/mL dexamethasone, 1.34 ng/mL recombinant human basic fibroblast growth factor analog (Fox et al., 1988), 10-5 M putrescine, 1 mg/mL bovine serum albumin (fraction V), and 50 pg/mL fetuin (Sigma) from fetal calves. (3) Production medium 2 (for the suspension culture process): DMEM/F12 supplemented with 2 mM glut-
@ 1994 American Chemical Society and American Instkute of Chemical Engineers
40
amine, 220 pg/mL serine, 65 pg/mL aspartate, 80 pg/mL asparagine, 250 pg/mL cysteine, 0.05 mM 2-mercaptoethanol, 10 pg/mL insulin, 10 pg/mL transferrin, 10 ng/ mL sodium selenite, 50 ng/mL dexamethasone, 2 pg/mL ethanolamine, 2.67 ng/mL recombinant human basic fibroblast growth factor analog, M putrescine, 1mg/ mL bovine serum albumin (fraction V), and 150 pg/mL fetuin I1 (Waitaki) derived from neonate calves. Cell Culture Conditions. Two different processes were used for producing r-HuTK by recombinant CHO cells. Sample TC018 was obtained using a microcarrier process, while samples TC020 and TC021 were obtained using a suspension cell recycle process. The cells were grown from identical vials of the same clone. There was no adaptation for the serum-free suspension culture. The final population density level for the cells in the three experiments were not determined. These two processes utilized different culture media as described here. (1)Microcarrier process: Cells were grown on Cytodex 2 microcarriers (Pharmacia) in a 2-L Holtz bioreactor. Cells [(4-5) X 109 were added to 3 g/L microcarriers. The cells were allowed to attach to the carriers and grow in amedium containing 5% fetal bovine serum (FBS) (see growth medium composition). After the microcarriers reached confluency, the medium was changed to a serum-free medium containing several supplements (see production medium 1). After a 2-day production period, the medium was harvested for purification. Fresh medium was added to the cells, and the harvest cycle was repeated several times. The pH was maintained at 7.0 f 0.2 and dissolved oxygen was maintained at 50% f 10% of air saturation. The cell concentration in the reactor averaged 3 X 106 cells/mL. (2) Serum-free suspension cell recycle process: The cells were grown in suspension in a serum-free medium supplemented with some proteins and other components (see production medium 2). In this case, the cells were grown in a 20-L reactor (Biolafitte ICC-20) with continuous medium perfusion. The cells were retained within the reactor by means of a recycle system which utilized a tangential flow filter (Millipore Prostak). The medium perfused through the microporous filter, while the cells were recycled back into the fermentor. The cell concentration in the reactor averaged 3 X lo6 cells/mL. Approximately 3 L of cell suspension was purged from the reactor every day to maintain the viability of the cells in the reactor at a high level. In this manner, the reactor was maintained at a steady state for approximately 3 months. The pH was maintained at 7.00 f 0.2 and dissolved oxygen was maintained at over 50% of air saturation. Conditioned media designated as TC020 and TC021 were harvested at 2 and 3 months, respectively. Sample Purification. Samples TC018, TC020, and TC021 were purified according to procedures as previously described (Lin and Lu, 1989). The culture medium containing expressed prokallikrein product was concentrated by diafiltration in 10 mM Tris-HC1 (pH 7.2) and then loaded onto a Q-Sepharose column equilibrated with the same buffer, as described. The prokallikrein fractions were eluted with a linear NaCl gradient from 0.05 to 0.5 M. The fractions containing partially purified prokallikrein were pooled, treated with thermolysin (1/200 w/w ratio) for 6 h at room temperature, and dialyzed against 10 mM NaOAc (pH 5.0) overnight at 4 "C. This protease treatment completely converted prokallikrein to biologically active kallikrein. The dialyzed kallikrein sample was subsequently purified by several steps of column chromatographicseparation,including DEAE-Sepharose
Biotechnol. Prog., 1994, Vol. 10, No. 1
chromatography, octyl-Sepharose chromatography, and S-200gel filtration. A final purified kallikrein gave a single diffused band in reducing SDS-PAGE gels and had enzymatic activity comparable to that of purified human urinary kallikrein. Analytical Isoelectric Focusing (IEF). IEF was carried out on a thin-layer polyacrylamide gel in the presence of urea using Servalyte 3-5 ampholines mixed with Servalyte 3-10. Coomassie Brilliant Blue G-250 (Serva, Heidelberg, Germany) was used for protein staining, and the stained gels were scanned using an UltroScan XL laser densitometer (LKB, Pharmacia) at 480 nm. Substituted Sialic Acids. Sialic acids were released under mild conditions following treatment with 0.5 mL of 2 M acetic acid for 3 h at 80 "C. The released sialic acids were derivatized with 1,2-diaminodimethoxybenzeneto form the highly fluorescent oxalidinone derivatives (Hara et al., 1986). The derivatized sialic acids were then separated by HPLC on a C18 reversed-phase column (25 X 0.46 cm) and detected by fluorescence. Desialylation of Oligosaccharides. For chromatographic experiments, oligosaccharides were desialylated by heating with 0.01 N HC1 at 80 "C for 60 min. The samples were cooled, neutralized with 0.01 N NaOH, and passed through the Bio-Rad P-2 column with distilled water as the eluant. For the IEF experiments, removal of sialic acids was performed by enzymatic digestion with 150 munits of neuraminidase in 35 pL of 50 mM sodium acetate buffer (pH 5.5) at 37 "C for 18 h. Anion-Exchange Chromatography. Low and high pH separations were carried out as described (Watson et al., 1992; Watson and Bhide, 1993) on a Dionex system, which consisted of a Dionex Bio-LC gradient pump, a CarboPac PA-1 column (analytical 4.6 X 250 cm or semiprep 9 X 250 mm), with pulsed amperometric detection. The following pulse potentials were used for detection: E = 0.00 V, E = 0.70 V, and E = -0.85 V. Low pH elution conditions used 2-100 mM sodium acetate over 60 min at pH 5.0. High pH elution was carried out using the same sodium acetate conditions in the presence of 100 mM sodium hydroxide. The flow rate for each elution was 1 mL/min. Fast-Atom Bombardment Mass Spectrometry (FAB-MS). FAB-MS was carried out on desialylated oligosaccharides as their ABEE derivatives (Webb et al., 1988). Individual oligosaccharides were placed in 5-mL Reacti-vials and dissolved in 10 pL of water. The reagent mixture, 40 pL, was added, and the vials were heated at 80 "C for 40 min. The sample was cooled, and 1mL each of H2O and CHzClz were added. The vials were vortexed and the water layer was removed. A second aliquot of 1 mL of water was added and the aqueous layers were combined. The aqueous extract was concentrated to approximately 0.5 mL, and the desialylated oligosaccharides were isolated using Biogel P-2 with water as the eluent. Results SDS-PAGE. SDS-PAGE determinations of all three r-HuTK samples showed essentially the same results, with the major band migrating with a molecular mass of approximately 40-43 kDa (results not shown). Isoelectric Focusing. Isoelectric focusing of the r-HuTK preparations TC018, TC020, and TC021 was carried out as described. All samples separated into many fractions, consistent with isoforms having p l values in the range of 3-5 (Figure 1). Samples TC020 and TC021 displayed the same number and relative amounts of
41
Bbtechnol. Rog., 1994, Vol. 10, No. 1
110
120
130
140
150
IB B
0.1
2
I
( 110
120
130
140
150
IC C
0.1
1 110
I
120
I
I
130
I
I
I
140
150
,
(DISTANCE mm)
Figure 1. Isoelectricfocusing separation,pH 3-5, of glycoforms of (A) TC018, (B) TC020,and (C)TC021. isoforms with identical migration patterns. In contrast, TC018 showed the isoforms with slightly higher plvalues. The relative proportions of the separated protein bands were obtained by integration as described. These profiles are commonly found in glycoproteins that contain different numbers of sialic acids due to the effect of the acidic carboxyl group, which has a pK, of 2.7. To show that these IEF differences were mainly due to sialic acid, each sample was digested with neuraminidase and again reanalyzed by IEF. In all three samples, all of the bands decreased to insignificant amounts, indicating that the microheterogeneity was due to the variable sialic acid content present in the protein bands (data not shown). Sialic Acid Profiles. The relative amounts of sialic acids present were determined as described. Following mild acid hydrolysis, released sialic acids were converted to fluorescent DMB derivatives and resolved by reversedphase HPLC. Figures 2A-C shows the sialic acid profiles obtained from TC018, TC020, and TC021, respectively. The amount of sample injected was adjusted such that the peak heights of N-acetylneuraminic acid were equal in all three samples in order to directly comparethe relative amounts of other sialic acids present. Peak 1is identified as N-glycolylneuraminic acid, and its level in all three samples is essentially the same at - 5 % . Peak 2 is N-acetylneuraminic acid, the major sialic acid present, while the remaining smaller peaks that elute after Nacetylneuraminic acid are tentatively identified as various acetylated N-acetylneuraminic acids. While N-glycolylneuraminic and N-acetylneuraminic acids could be identified and determined using commercially available standards, no such standards were available for any of the other substituted sialic acids, and their tentative identification as acetylated sialic acids is proposed from the relative retention times reported by others (Hara et al., 1989).
1 -L
A
Comparison of N-Linked Oligosaccharide Profilm. Released oligosaccharides from TC018, TC020, and TC021 were chromatographed under low pH anionexchange conditions and exhibited markedly different elution profiles (Figure 3A-C). The oligosaccharide fractions separated into multiple peaks that contained one, two, three, and four sialic acids. Determination of the number of sialic acids and tentative identification of the oligosaccharides were established from chromatography of monosialyl diantennary, disialyl diantennary, trisialyl triantennary, and tetrasialyl tetraantennary oligosaccharides obtained from Oxford Glycosystems and Dionex, the glycoproteins chorionic gonadotropin, fetuin, and a-acid glycoprotein, and the recombinant glycoproteins platelet-derived growth fact or and erythropoietin, as described (Watson et al., 1992;Watson and Bhide, 1993; Watson et al., 1993). No significant differences were apparent in the mapping profiles between samplesTC020 and TC021, and both contained the same relative amounts of all of the various mono-, di-, tri-, and tetrasialylated oligosaccharides. In contrast, TC018 contained relatively smaller amounts of the tetrasialyl species, and there was a shift to oligosaccharide populations with lower numbers of sialic acids. SamplesTC018, TC020, and TC021 were hydrolyzed under mild conditions to remove sialic acid and rechromatographed. Each sample showed a single major peak having the same retention time as that from
’
Biotechnol. Prog., 1994,Vol. 10, No. 1
-
A
Wonosialyl
Disiaiyl
Trisialyl
Tetrasialyl
2
I,
10
15
20
25
30
35
40
, , , ,I
, , /
45
50
55
J
60
'
'
'
I
'
'
~ ' ~ " ' ' ~ ~ ~ " ' " ' ' ' ' ' ~ 20 25 30 35 40
l
15
10
--
3
- d
5
TIME (Min.)
TIME (Min.)
TetrasialyI
Trisialyl
18
Monosial I
4 -D i s W
0
5
10
15
20
IC
25 30 35 TIME (Min.)
40
45
50
55
60
I
-Trisialyl
5
10
,
,
,
,
15
I
I
,
,
.
I
20
,
,
,
1
1
,
1
~
~
30
25
I
~
I
I
~
40
35
TIME (Min.)
Tetrasialyl
C
3
5 TIME (Min.)
Figure 3. Low pH anion-exchange separation of reduced sialyloligosaccharidesfrom kallikrein samples (A) TC018,(B)
TC020,and (C)TC021.
sialic acid (chromatogram not shown), indicating that all peaks present contained sialic acid. Next, a comparison of the glycan profile after desialylation was carried out. Chromatography was carried out under high pH anion-exchange conditions in order to resolve the desialylated species. The results are shown in Figure 4A-C. Peaks 1-5 were identified from their chromatographic retention times before and after desialylation and after spiking with known oligosaccharides that had been isolated and characterized from recombinant human erythropoietin (Watson et ai., 1993). Chromatographic profiling of desialylated oligosaccharidesprovides an alternate mapping technique, but one that is less diagnostic of the total glycosylation heterogeneity than can be obtained from low pH chromatography of sialylated species. Under high pH eluant conditions, the separation of oligosaccharides occurs in the order of increasing size, in agreement with previous reports (Basa and Spellman, 1990). Comparison of the chromatographic profiles, Figures 4A-C, reveals a significant difference in the types of antennary oligosaccharides present. Sample TC018 contains more biantennary complex type oligosaccharides, while samples TC020 and TC021 have much higher amounts of tetraantennary types, with zero, one, and two N-acetyllactosamine complex type oligosaccharides. Fastatom bombardment mass spectrometry of the ABEE derivatives of the unreduced isolated oligosaccharidesgave (M + Na)+ ions at mlz values of 1959, 2324,2689, 3054, and 3419. These values correspond to the derivatized diantennary, triantennary, and tetraantennary (oligosaccharides, and tetraantennary oligosaccharidesbearing one and two N-acetyllactosamine extensions (figures not shown), respectively.
l , , , , l , , , , l , 1,
5
10
15
20
I , ,
25
,
I
30
/
\
t
.
l
35
t
,
/ I
40
TIME (Min.)
Figure 4. High pH anion-exchangeseparation of desialylated oligosaccharidesfrom (A)TC018,(B)TC020,and (C)TC021. Assignment of peaks 1-5 is described in the text. Complete structural characterization of all of the oligosaccharides is underway and will be reported when these studies are complete. Preliminary results are in agreement with the identities of the oligosaccharides reported here.
Discussion The environment in which cells are grown plays a crucial role with respect to the assembly of oligosaccharide structures on glycoproteins. Culture medium components, including growth factors and proteins, along with glucose concentration and ammonium ion concentration (Gooche and Monica, 1990) can influence oligosaccharide heterogeneity. The mechanistic cause for these differences is frequently unknown, but the consequences of such processing have major implications in the production of glycoproteins intended for therapeutic use in human subjects. The type of cell culture production system used for the scaleup of glycoproteins may also influence oligosaccharide heterogeneity. Large-scale production of glycoproteins must often uses immobilized cell culture processes, such as microcarriers. Microcarrier culture involves the growing and maintaining of attached cells on small beads in a stirred vessel. The advantages of microcarrier systems include the following: increased productivity due to higher cell concentration, ease of medium replenishment, subculturing, and ease of product recovery from cells. In the present study, there were significant differences in glycosylation patterns, most notably in the decreased sialylation in human kallikrein expressed by cells cultured on microcarrier beads as compared to the suspension culture.
I
I
I
~
~
i
43
Bbtechnoi. Rog., 1994, Vol. 10, No. 1
There were also some differencesin the glycan structures of human kallikrein between the microcarrier and suspension culture systems, with the latter having a higher proportion of more complex oligosaccharides. These were much less significant in comparison to the differences in sialylation patterns. N-Acetylneuraminic acid was the major sialic acid in each sample, and all three samples contained small amounts of N-glycolylneuraminic acid ( < 5 % ) . Samples removed a t times of 2 and 3 months from the suspension cell recycle process showed no difference in glycosylation heterogeneity, indicating that glycosylation had reached a steady-state condition by 2 months and that no further changes occurred in production and glycosylationheterogeneity by continuing the process for any longer periods of time. Samples TC018, TC020, and TC021 were all purified by the exact same procedure, thereby eliminating the possibility that the observed differencesin glycosylation patterns were caused by factors other than differences in the culture processes. The results shown here illustrate the need to examine glycosylationpatterns in recombinant glycoproteins when cell culture production methods are under consideration for large-scaleproduction. The recent introduction of high pH anion-exchange chromatography coupled with pulsed amperometric detection has proven a highly versatile and useful technique for determining carbohydrates. A series of reports has described its application for the analysis of monosaccharides (Hardy and Townsend, 1988), high mannose (Basa and Spellman, 1990), and acidic oligosaccharides (Townsend et al., 1989). Its subsequent application as a mapping technique for complex carohydrates released from recombinant-derived glycoproteinshas been described in several reports (Spellman et al., 1989; Kumarasamy, 1990; Anumula and Taylor, 1991; Hermentin et al., 1992). The rationale for high pH conditions in separating oligosaccharides is based on the ionization of hydroxyl groups to form oxyanions, which can then undergo chromatography on anion-exchange columns. In two recent reports, we have reexamined this proposal and have shown that low pH conditions result in superior separations of N-linked sialic acid oligosaccharides obtained from recombinant glycoproteins (Watson et al., 1992; Watson and Bhide, 1993). Since these oligosaccharides eluted in the same order under both high and low pH conditions, we concluded that oxyanion formation is not the sole reason for the observed separation under high pH conditions and that other mechanisms are involved. Use of the low pH chromatographic system as described here providedkhe most diagnostic information on the types of oligosaccharides present, as well as a means to identify the nature of the differences. Much less information was obtained from IEF separation of glycoproteins,while SDSPAGE does not yield information about changes in glycosylation pattern. The exact reason for the observed differences in the glycosylation of human recombinant kallikrein has not been determined; however, there are several possibilities. Samples TC020 and TC021 produced from cells in suspension culture contain more sialylated chains with highly branched structures and N-acetyllactosamine groups, which take longer to synthesize by cells, and the glycosylation differences shown here may reflect the different production or secretion rates of r-HuTK by the cells in two different culture conditions. In many instances, cell monolayers covering microcarrier beads overgrow to form multiple layers of cells, limiting access to medium components and release of waste product, which are necessary for proper cellular physiology. Thus, these
suboptimal conditions can lead to changes in oligosaccharide processing. It is also plausible that differences in culture medium composition contributed to the determined alterations in glycosylation. It is also possible that the differences in glycosylation observed could be due to differences in the rate of kallikrein synthesis per cell under the two different culture conditions and/or carbohydrate synthesis. However, regardless of the exact reasons for the observed differences in glycosylation that are occurring with rHuTK, this article demonstrates that significant differences in N-linked glycosylation can occur with different culture conditions and that these differences may be readily determined using the anion-exchange chromatography described here. At the present time, it is not entirely clear why these differences in glycosylation, specifidly sialylation, are occurring. Studies to further determine possible reasons are underway. Acknowledgment We thank Dr. Terry Lee (Beckman ResearchsInstitute of the City of Hope) for FAB-MS analyses, Blair McNeill and Steve Ogden for their assistance in microcarrier and suspension culture production, and Dr. Larry Tsai for helpful discussions on cell culture processes and also for reviewing the manuscript. We also thank Ms. Joan Bennett for her help in the preparation of this manuscript. Literature Cited Anumula, K. R.; Taylor, P. B. Rapid Characterization of asparagine-linkedoligosaccharidesisolated from glycoproteins using a carbohydrate analyzer. Eur. J. Biochem. 1991,195, 269-280.
Basa, L. J.; Spellman, M. W. Analysis of glycoprotein-derived oligosaccharidesby high-pH anion-exchangechromatography. J. Chromatogr. 1990,499, 205-220. Erdos, E. G.; Wilde, A. F. Handbook of Experimental Pharmacology of Bradykinin, Kallidin, and Kallikrein, Supplement. In Handbook of Experimental Pharmacology;Springer: Berlin, 1979, Vol. 25. Fiedler, F.; Fink, E.; Rscheche, H.; Fritz, H. Porcine glandular kallikreins. Methods Enzymol. 1981,80,493-532. Fox, G. M.; Schiffer, S. G.; Rohde, M. F.; Tsai, L. B.; Banks, A. R.; Arakawa, T. Production, biological activity, and structure of recombinant basic fibroblast growth factor and an analog with cysteine replaced by serine. J. Biol. Chem. 1988, 263, 18452-18458.
Goochee, C. F.; Monica, T. Environmental effects on protein glycosylation. BiolTechnology 1990,8, 421-426. Hara, S.; Hamaguchi, M.; Takemori, Y.; Nakamura, M. J. Highly sensitive determination of N-acetyl- and N-glycolylneuraminic acids in human serum and urine and rat serum by reversedphase liquid chromatography with fluorescence detection. J. Chromatogr. 1986, 377, 111-119. Hara, S.; Yamaguchi,M.; Takemori, Y.; Furuhata, K.; Ogura, H.; Nakamura, M. Determination of mono-0-acetylated N-acetylneuraminic acids in human and rat sera by fluorometric highperformance liquid chromatography. Anal. Biochem. 1989, 179, 162-166.
Hardy, M. R.; Townsend, R. R.; Lee, Y. C. Monosaccharide analysisof glycoconjugatesby anion-exchangechromatography with pulsed amperometric detection. Anal. Biochem. 1988, 170, 54-62.
Hermentin, P.; Witzel, R.; Vliegenthart, J. F. G.; Kamerling, J. P.; Nimitz, M.; Conradt, H. S. A strategy for the mapping of N-glycans by high-pH anion-exchange chromatography with pulsed amperometric detection. AnaE. Biochem. 1992, 203, 281-289.
Biotechnol. Prog., 1994, Vol. 10, No. 1
44
Irie, A.; Kuehiro, H.; Kodama, J.; Ohata,M.; Miyake,Y. In Recent Progress on Kinins; Friz, H., Dietze, G., Fiedler, F., Haberland, G. F., Eds.; Birkhauser Verlag: Basel, 1982;pp 131-136. Kellerman, J.; Lottapeich, F.; Geiger, R.; Deutzmann, R. Human urinary kallikrein-amino acid sequence and carbohydrate attachment sites. Protein Seq. Data Anal. 1988,1,177-182. Kumarasmy, R. Oligosaccharide mapping of the therapeutic glycoproteins by high-pH anion-exchange high-performance liquid chromatography. J. Chromatogr. 1990,512,149-155. Lin, F.-K.; Lu, H. S. European Patent Application No. 0297913, 1989. Lu, H. S.; Lin, F.-K.; Chao, L.; Chao, J. Human urinary kallikrein-complete amino acid sequence and sites of glycosylation. Int. J. Pept. Protein Res. 1989,33, 237-249. Pisano, J. J.; Austen, K. F. DHEW Publ. (NIH) (U.S.) NIH 76791 (1976). Spell", M. W.; Basa, L. J.; Leonard, C. K.; Chakel, J. A.; OConnor, J. V.; Wilson, S.; van Halbeek, H. Carbohydrate structures of human tissue plasminogen activator expressed in Chinesehamster ovarycells. J.Biol. Chem. 1989,264,1410014111. Takahashi, S.; Irie, A.; Miyake, Y. Primary structure of human urinary prokallikrein. J. Biochem. 1988,104,22-29. Townsend, R.R.;Hardy, M. R., Cumming, D. A,; Carver, J. P.; Bendiak, B. Separation of branched sialylated oligosaccharides
using high pH anion-exchange chromatography with pulsed amperometric detection. Anal. Biochem. 1989,182,1-8. Watson, E.;Bhide, A. Carbohydrate analysis of recombinantderived erythropoietin. LC-GC 1993,11, 216-220. Watson, E.; Bhide, A.; Kenney, W. C.; Lin, F. K. Highperformance anion-exchange chromatography of asparaginelinked oligosaccharides. Anal. Biochem. 1992,205,9G95. Watson, E.; Bhide, A.; van Halbeek, H. Structure determination of the major sialylated oligosaccharidechains of recombinant derived erythropoietin expressed in Chinese hamster ovary cells. Presented at the 41st ASMS Conference on Mass Spectrometry and Allied Topics, San Francisco, CA, May 30June 4,1993. Webb, J. W.; Jiang, K.; Gillece-Castro, B. L.; Tarentino, A. L.; Plummer, T. H.; Byrd, J. C.; Fisher, S. J.; Burlingame, A. L. Structural characterizationof intact, branched oligosaccharides by high performance liquid chromatography and liquid secondary ion mass spectrometry. Anal. Biochem. 1988,169, 337-349. Accepted September 13,1993.' e Abstract
1993.
published in Aduance ACS Abstracts, December 1,
