Glycobiology Advance Access originally published online on May 22, 2006
Glycobiology 2006 16(9):833-843; doi:10.1093/glycob/cwl004
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
GlycoPEGylation of recombinant therapeutic proteins produced in Escherichia coli
3 Neose Technologies, Inc., 102 Witmer Road Drive, Horsham, PA 19044; 4 Neose Technologies, Inc., 6330 Nancy Ridge Road, Suites 102–103, San Diego, CA 92121; and 5 Department of Medical Biochemistry and Genetics, Glycobiology, University of Copenhagen, Blegdamsvej 3, DK-2200 Copenhagen N, Denmark
1 To whom correspondence should be addressed; e-mail: dzopf{at}neose.com
2 Present address: Noordwijkerhout 2211 AR, Schaepmanlaan 19, The Netherlands
Received on February 8, 2006; revised on May 11, 2006; accepted on May 13, 2006
 |
Abstract |
|---|
 Top Abstract IntroductionResultsDiscussionMaterials and MethodsSupplementary DataConflict of interest statementAcknowledgmentsReferences |
|---|
Covalent attachment of polyethylene glycol, PEGylation, has been shown to prolong the half-life and enhance the pharmacodynamics of therapeutic proteins. Current methods for PEGylation, which rely on chemical conjugation through reactive groups on amino acids, often generate isoforms in which PEG is attached at sites that interfere with bioactivity. Here, we present a novel strategy for site-directed PEGylation using glycosyltransferases to attach PEG to O-glycans. The process involves enzymatic GalNAc glycosylation at specific serine and threonine residues in proteins expressed without glycosylation in Escherichia coli, followed by enzymatic transfer of sialic acid conjugated with PEG to the introduced GalNAc residues. The strategy was applied to three therapeutic polypeptides, granulocyte colony stimulating factor (G-CSF), interferon-alpha2b (IFN-
2b), and granulocyte/macrophage colony stimulating factor (GM-CSF), which are currently in clinical use.
Key words:
G-CSF
/
glycosylation
/
GM-CSF
/
IFN-2b
/
PEGylation
|
Introduction |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and MethodsSupplementary DataConflict of interest statementAcknowledgmentsReferences |
|---|
Modification of protein drugs by covalent attachment of polyethylene glycol (PEG) can improve physical and thermal stabilities, protect against degradation by enzymes, enhance solubility, prolong circulating half-life, and, in some cases, reduce immunogenicity of the polypeptide (Katre, 1993
; Francis et al., 1998; Harris et al., 2001; Harris and Chess, 2003). The circulatory half-life of a parenterally administered therapeutic protein may be extended significantly by covalent attachment of one or more chains of PEG. For example, conjugation of 20 kDa PEG to the amino terminus of Escherichia coli-produced G-CSF (Filgrastim; Amgen, Thousand Oaks, CA) gives a product (Pegfilgrastim; Amgen) whose plasma half-life in humans is increased by a factor of 
20 (Zamboni, 2003; Neupogen, 2005). PEGylation of proteins represents a significant manufacturing challenge in that coupling reactions between chemically activated PEG and free amino groups, even after extensive optimization, commonly result in multiple PEGylated isoforms with nonuniform chemical and pharmaceutical properties (Kinstler et al., 1996; Francis et al., 1998; Foser et al., 2003; Jolling et al., 2004).
Drug activity typically relies upon contact of a properly folded active site of a protein agonist with the complementary region of its cognate receptor. In the past, considerable effort has gone into tailoring a PEG coupling strategy for each individual protein drug to avoid modification of its active site (Harris and Chess, 2003). Chemically activated PEGs that react with lysyl or histidyl residues have generated successful long-acting products from a subset of drug proteins, but reactivities of target substrates vary in response to local chemical environments (e.g., charge) created by neighboring amino acids at the surface of the folded polypeptide, and for many proteins, alternative approaches have proved necessary (Katre, 1993; Harris and Chess, 2003). These have included altering amino acid sequences by the introduction of an unpaired cysteine that can be selectively PEGylated (Yang et al., 2003), random mutagenesis and high-throughput screening to select bioactive muteins lacking lysyl groups in the region of the active site (Yoshioka et al., 2004), and the use of the enzyme transglutaminase to introduce alkyl PEG on to glutamine residues (Sato, 2002).
We have designed a novel strategy for site-directed enzymatic PEGylation based on the finding that sialic acid covalently substituted with PEG at the 5'-amino position can be enzymatically transferred to glycan acceptors on glycoproteins. Several clinically important drugs such as granulocyte colony stimulating factor (G-CSF), interferon-alpha2b (IFN-2b), and granulocyte/macrophage colony stimulating factor (GM-CSF) are naturally O-glycosylated human glycoproteins but are manufactured by recombinant expression in E. coli as nonglycosylated polypeptides. In the work described here, we utilized a simple two-step process for site-directed PEGylation using enzymatic GalNAc O-glycosylation followed by enzymatic PEGylation of the introduced O-glycans (Figure 1). Escherichia coli-expressed G-CSF, IFN-2b, and GM-CSF proteins were selectively GalNAc O-glycosylated at their natural O-glycosylation sites using specific recombinant human UDP-GalNAc: polypeptide N-acetylgalactosaminyltransferase (GalNAc-T) isoforms. The selection of the appropriate GalNAc-transferase isoform was based on a simple screening assay of substrate specificities with short synthetic peptides covering O-glycan acceptor sites of interest. Enzymatic PEGylation of the introduced GalNAc O-glycan was achieved with sialyltransferases that can transfer sialic acid conjugated with PEG. The process forms a biologically active, chemically homogeneous GlycoPEGylated protein with extended plasma half-life.
|
) by a polypeptide GalNAc transferase, for example, GalNAc-T2. Once incorporated, the GalNAc residue may serve as a direct acceptor for PEGylated sialic acid transferred by ST6GalNAc-I. (B) Polypeptides such as GM-CSF may contain multiple closely positioned natural nonutilized O-glycosylation sites (orange) in which a serine (S) or threonine (T) residue may serve as an acceptor for selective addition of GalNAc (
) by Core1 GalT, the product of which creates an acceptor for transfer of PEGylated sialic acid (
-PEG) by ST3Gal-1. The enzyme reactions can be run separately with intermediate purification of the product or in a single vessel without intermediate purification steps. |
Results |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and MethodsSupplementary DataConflict of interest statementAcknowledgmentsReferences |
|---|
Selection strategy for GalNAc-T isoforms for site-directed O-glycosylation
GalNAc-T isoforms have unique and partly overlapping acceptor substrate specificities. Although distinct consensus acceptor sequence motifs have not been elucidated, studies indicate that the primary sequence context is the major determining factor for substrate specificities (Hassan et al., 2000
). Thus, screening short 15–20mer peptide sequences derived from a protein of interest by in vitro glycosylation assays with GalNAc-T isoforms is predicted to provide reasonable assessment of GalNAc-Ts function with the protein. We used short synthetic peptides covering natural O-glycosylation sites in G-CSF (Figure 2A), IFN-2b (Figure 3A), and GM-CSF (Figure 4A) to evaluate GalNAc-T isoforms suitable for site-directed GalNAc glycosylation of the E. coli-produced unglycosylated cytokines. Monitoring reactions by matrix-assisted laser desorption ionization-time-of-flight mass spectrometry (MALDI-TOF/MS) time-course analysis revealed that the glycosylation of G-CSF and IFN-2b peptides were essentially complete, while the glycosylation of the GM-CSF peptide did not go to completion with two and three residues of GalNAc incorporated. Native G-CSF and IFN-2b are O-glycosylated at a single site (Thr133 and Thr106, respectively), while GM-CSF is O-glycosylated at multiple sites (Ser5, Ser7, Ser9, and/or Thr10) (Kaushansky et al., 1987, 1992; Forno et al., 2004). Recombinant forms of these cytokines expressed in mammalian cells are similarly O-glycosylated, but when expressed in E. coli, they are not. Among the six human recombinant GalNAc-Ts tested, GalNAc-T2 was identified as the only candidate isoform for glycosylation of G-CSF (Figure 2A) and GM-CSF (Figure 4A), while several isoforms including T2 were candidates for glycosylation of IFN-2b (Figure 3A).
|
|
|
GlycoPEGylation of E. coli-expressed cytokines
Site-specific GalNAc glycosylation of the reported natural O-glycosylation sites in G-CSF, IFN-2b, and GM-CSF recombinant proteins was obtained as predicted from the screening assay with the GalNAc-T2 isoform. MALDI-TOF/MS analysis suggested incorporation of a single-GalNAc residue into G-CSF and IFN-2b (Figures 2B and 3B), which was further confirmed by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) analysis (Figures 2C and 3C). Peptide mapping of a GluC plus trypsin digest of G-CSF(GalNAc) by liquid chromatography – mass spectrometry/mass spectrometrometry (LC-MS/MS) confirmed the presence of a doubly charged peptide ion 1282.27, representing 124LGMAPALQPT133QGAMPAFASAFQR146 plus GalNAc, in which GalNAc was shown by MS/MS peptide sequencing to be linked to T133 (data not shown). Similarly, a doubly charged ion from GalNAc-T2-glycosylated IFN-2b (1018.69, representing 97ACVIQGVGVT106ETPLMKE113 plus GalNAc) was shown by MS/MS peptide sequencing to be glycosylated at T106. Ions for the corresponding unglycosylated peptides were not observed. GalNAc-T2 incorporated two GalNAc residues into GM-CSF at Ser7 and Ser9, and a third site of low incorporation was not identified.
Sialyltransferases including ST6GalNAc-I were initially found to transfer cytidine monophosphate (CMP)-sialic acid with 20 kDa PEG linked through the 5'-amino nitrogen of the sialic acid residue to simple acceptor substrates (C. Bowe, et al., manuscript in preparation). Here, we used this feature of ST6GalNAc-I for efficient transfer of sialic acid-PEG from CMP-sialic acid-PEG-20K to the enzymatically introduced GalNAc residues in G-CSF(GalNAc) and IFN-2b(GalNAc). The reaction proceeded essentially to completion, and the final purified products appear as single bands by SDS–PAGE analysis (Figures 2C and 3C). GlycoPEGylation of GM-CSF followed a modified strategy and did not result in similar homogenous glycosylation products (Figure 1B). GalNAc-T2 incorporated two to three GalNAc residues into the short synthetic peptide after prolonged incubation (22 h), and two residues with traces of one and three residues were incorporated into the GM-CSF protein (Figure 4A and B). A doubly charged ion from a peptide digest of the glycosylated GM-CSF protein product (1235.00, representing 4RS5PS7PS9T10QPWEHVNAIQE21 plus two GalNAc residues) was identified, and Edman degradation was used to determine that the major residues substituted with O-GalNAc were Ser7 and Ser9 (data not shown). Thus, the combined data from MS/MS peptide sequencing and chemical sequencing by Edman degradation indicate that in vitro O-glycosylation by GalNAc-T2 attaches GalNAc residues almost exclusively to Ser7 and Ser9, with trace incorporation after prolonged incubation into a third site which we have not identified. Preliminary experiments showed that ST6GalNAc-I could add only one mole of SiaPEG-20K to GM-CSF(GalNAc)2, presumably because once added, a bulky PEG substituent at one site sterically hinders approach of the enzyme to an acceptor GalNAc linked to a nearby amino acid. We hypothesized that addition of a sugar residue "spacer" to each GalNAc might provide a more favorable substrate for the addition of two SiaPEG-20K residues and found that at least 75% of polypeptide molecules can be converted to the di-PEGylated form using this approach. PEGylation of the two O-linked GalNAc residues occupying adjacent amino acid residues on GM-CSF(GalNAc)2 was therefore achieved via a different pathway, employing two enzymes in a "one-pot" reaction: Core1 GalT is known to add galactose in ß-linkage to the 3-O-position of GalNAc (Ju et al., 2002), and ST3Gal-1 is known to transfer sialic acid in -linkage to the 3-O-position of galactose linked ß1-3 to GalNAc (Blixt et al., 2002; Jeanneau et al., 2004). When combined in a single reaction mixture, the enzymes act consecutively such that addition of the first SiaPEG-20K proceeds essentially to completion, whereas addition of a second SiaPEG-20K proceeds to only 75% (Figure 4C, peak 2). After chromatographic separation, SDS–PAGE analysis shows that both the singly and doubly PEGylated products migrate as single bands with only a faint trace of doubly and triply PEGylated products, respectively (Figure 4D). A trace of triply PEGylated product was detected in some preparations (Figure 4D, lane 2 and Figure 4C, peak 1). Like the addition of SiaPEG-20K by ST6GalNAc-I, the addition of Gal using Core1 GalT, followed by SiaPEG-20K using ST3Gal-1 proceeds under mild buffer conditions, and its selectivity for the O-linked GalNAc carbohydrate acceptor assures a well-defined product.
Pharmacokinetics of GlycoPEGylated G-CSF and IFN-
Comparative analysis of plasma concentrations of G-CSF (unmodified E. coli-produced protein), G-CSF(GalNAc-SiaPEG-20K), and G-CSF(PEG-20K) (E. coli-produced G-CSF with 20 kDa PEG chemically coupled to the amino terminus) at various time points following bolus intravenous administration in rats shows that the clearance of GlycoPEGylated G-CSF(GalNAc-SiaPEG-20K) similar to G-CSF(PEG-20K) was reduced more than 10-fold compared with G-CSF (Figure 2D and Table I). The terminal phase plasma half-life observed for G-CSF(GalNAc-SiaPEG-20K) was 8.78 h, a value comparable with the t1/2 for G-CSF(PEG-20K) (7.74 h), but increased more than 5-fold over t1/2 for G-CSF (1.71 h). The calculated area under the curve for G-CSF(GalNAc-SiaPEG-20K) is increased 13-fold with respect to G-CSF and 1.8-fold with respect to G-CSF(PEG-20K). This data indicate that the pharmacokinetic profile of G-CSF(GalNAc-SiaPEG-20K) is comparable with that of a chemically PEGylated G-CSF molecule in current clinical use and may provide a slightly improved prolonged systemic exposure following a single dose. GlycoPEGylated IFN-2b(GalNAc-SiaPEG-20K) also showed a significantly slower clearance rate compared with the unmodified IFN-2b protein produced in E. coli (Figure 3D). Precise calculation of pharmacokinetic parameters for radiolabeled protein is compromised, however, by uncertainties regarding metabolic turnover of the [125I]-labeled protein more than 18 h after intravenous administration.
|
Bioactivity of G-CSF and IFN-
Comparative analysis of binding affinities of PEGylated and nonPEGylated G-CSF variants for the G-CSF receptor demonstrated that receptor-binding properties of the molecules were largely preserved: Ki values for G-CSF(GalNAc-SiaPEG-20K), G-CSF(PEG-20K), and G-CSF were 1.14, 3.2, and 0.4 nM, respectively. Specific activities of the three compounds as stimulators of in vitro proliferation of NFS-60 murine myeloid leukemia cells were 1.4 x 104, 3.8 x 104, and 9.1 x 104 U/µg, respectively, which is comparable with the reported value for rhG-CSF of 105 U/µg (Neupogen, 2005). When injected as a single intravenous dose in mice, G-CSF(GalNAc-SiaPEG-20K) caused the peripheral blood white blood cell (WBC) count to increase to a maximum 5.6-fold above baseline after 48 h with return to baseline at 84 h, a response profile comparable with, or somewhat better than, that of G-CSF(PEG-20K) (Figure 2E). By contrast, the same dose of unmodified E. coli-produced G-CSF raised the WBC count only 2.8-fold at 24 h, with return to baseline at 48 h. The ED50 of IFN-2b(GalNAc-SiaPEG-20K) in the Madin–Darby bovine kidney (MDBK)-vesicular stomatitis virus (VSV) assay was 16.8 pg/mL compared with 439 pg/mL for IFN-2b (Figure 3E).
|
Discussion |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and MethodsSupplementary DataConflict of interest statementAcknowledgmentsReferences |
|---|
This study demonstrates an alternative strategy for site-directed attachment of PEG chains by the enzymatic process termed GlycoPEGylation. Feasibility of the strategy was confirmed with two therapeutic proteins, G-CSF and IFN-
2b, which are currently in clinical use. The strategy relies on enzymatic transfer of GalNAc to natural O-glycosylation sites that are not glycosylated in the recombinant proteins expressed in E. coli. Since both G-CSF and IFN-2b proteins retain their respective biological activities when expressed in E. coli as nonglycosylated polypeptides (Bodo and Maurer-Fogy, 1985a, 1985b; Rotondaro et al., 1999), glycosylation is not required for folding or receptor binding. Steric maps constructed from X-ray crystal data show that the glycosylation sites of these cytokines are remote from their active sites (Radhakrishnan et al., 1996; Aritomi et al., 1999). We, therefore, hypothesized that the attachment of O-glycans extended with PEG to the natural O-glycosylation sites would not impair their biological activities. Studies in an animal model that has generally proved predictive of myelogenic responses to G-CSF in humans (Lord et al., 2001) demonstrated that G-CSF(GalNAc-SiaPEG-20K) has a pharmacodynamic profile consistent with prolonged ability to stimulate proliferation and differentiation of stem cells to granulocytes. PEGylation of G-CSF has been proposed both to reduce renal clearance (Tanaka and Tokiwa, 1990) and to decrease cognate receptor affinity, thereby reducing clearance via receptor-mediated endocytosis (Kuwabara et al., 1995). Similarly, IFN-2b (GalNAc-SiaPEG-20K) exhibits antiviral titers comparable with the specific activities reported for chemically PEGylated interferons approved for human use (Youngster et al., 2002; Foser et al., 2003).
The GlycoPEGylation strategy relies on two unique features of glycosyltransferases: the unique substrate specificities of polypeptide GalNAc-T isoforms and the ability of glycosyltransferases to use modified donor nucleotide substrates. The GalNAc-T family contains up to 20 isoforms that transfer GalNAc to serine and/or threonine residues (Hassan et al., 2000; Ten Hagen et al., 2003), and this study demonstrates that the unique substrate specificity of these enzymes can be used for site-directed enzymatic O-glycosylation of proteins. The applied screening strategy using a few synthetic peptides that include natural O-glycosylation sites from a protein of interest rapidly identifies one or more isoforms useful for O-glycosylation at the corresponding natural site in the complete protein. These results provide further support for the importance of the primary sequence context in acceptor substrate specificities of GalNAc-Ts (Hassan et al., 2000). Introduction of O-linked GalNAc residues in proteins provides an acceptor site for several other glycosyltransferases that can add sialic acids, Gal, or GlcNAc monosaccharides.
In recent years, scaled-up production of recombinant glycosyltransferases overexpressed in mammalian or fungal host cell systems has provided soluble enzyme reagents suitable for industrial in vitro remodeling of glycan chains attached to pharmaceutically active glycoproteins. For example, the combined use of recombinant ST3Gal-III and FT-VI to link sialic acid and fucose, respectively, to create Slex epitopes at the termini of biantennary N-linked glycans on the human complement inhibitor sCR1 was carried out at the 10 g scale (Thomas et al., 2004). The product exhibited improved pharmacokinetic properties and a 10-fold increase in affinity for E-selectin. Similarly, pilot scale galactosylation of N-glycans of human immunoglobulin (IgG) to create molecules with >98% G2 glycoform was performed using recombinant ß4Gal-T1 in a reaction containing 1 kg of IgG-acceptor glycoprotein (Warnock et al., 2005). As these enzymes are required in relatively small amounts relative to the glycoprotein protein drug (0.1% on a mass basis), their contribution to overall cost of manufacturing may be offset by the value of enhanced performance characteristics of the product.
Several reports by others have previously demonstrated that glycosyltransferases may catalyze transfer of sugars with substituents. These include sialyltransferase ST6Gal-1 catalyzed transfer of sialic acids substituted at the C-9 position (Gross et al., 1989) and milk fucosyltransferase catalyzed transfer of fucose substituted at the C-6 position (Tsuboi et al., 2000). Here, we have employed the sialyltransferase ST6GalNAc-I, an enzyme, which naturally transfers sialic acid from the sugar nucleotide donor CMP-sialic acid onto the 6-O-position of O-linked GalNAc (Kurosawa et al., 2000). We have shown that this sialyltransferase can efficiently utilize a modified sugar nucleotide donor in which sialic acid is substituted at the 5'-amino position with a 20 kDa linear PEG chain, a substituent considerably larger than any previously described. Transfer occurs exclusively to single-GalNAc residues introduced in G-CSF and IFN-2b, giving a chemically uniform product with well-preserved bioactivity. A slightly modified pathway for GlycoPEGylation of GM-CSF was developed. GalNAc-T2 introduced two GalNAc residues in close proximity (Ser7 and Ser9), although the products were more heterogeneous than the products for G-CSF and IFN-2b. Since preliminary studies showed that the transfer of SiaPEG-20K by ST6GalNAc-I was incomplete, galactosylation to form the core 1 structure Galß1-3GalNAc before transfer of SiaPEG-20K with a different sialyltransferase, ST3Gal-1, was used.
In conclusion, we demonstrate a novel approach for site-directed enzymatic attachment of large PEG groups to the recombinant therapeutic proteins G-CSF, IFN-2b, and GM-CSF. Selective addition of sialic acid-PEG to O-linked GalNAc on a protein provides a novel, highly site-selective mechanism for PEGylation, enabling the manufacture of long-acting protein drugs with greater structural homogeneity as compared with PEGylated proteins prepared by conventional chemical methods (Roberts et al., 2002).
|
Materials and Methods |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and MethodsSupplementary DataConflict of interest statementAcknowledgmentsReferences |
|---|
Materials
CMP-sialic acid with 20 kDa linear methoxypolyethylene glycol linked to the 5'-amino nitrogen of the sialic acid residue (CMP-SiaPEG-20K) was prepared at Neose Technologies, Inc., (Horsham, PA; C. Bowe, et al., manuscript in preparation; DeFrees et al., 2004
). Filgrastim and Pegfilgrastim are products of Amgen (purchased from Besse Pharmaceutical Supply). GM-CSF and IFN-2b were purchased from Cell Sciences, Canton, MA. ITC-DTPA-Europium labeling kits were obtained from Perkin Elmer, Wellesley, MA, Catalog No. AD-0021.
Glycosyltransferase enzymes
Soluble secreted human polypeptide GalNAc transferases were expressed in insect cells and purified as described previously for GalNAc-T1, GalNAc-T2, and GalNAc-T3 (Wandall et al., 1997), GalNAc-T4 (Bennett et al., 1998), GalNAc-T6 (Bennett et al., 1999), and GalNAc-T11 (Schwientek et al., 2002). Soluble secreted chicken sialyltransferase, chST6GalNAc-I (GenBank accession No. X74946
[GenBank]
), was expressed in insect cells and purified to homogeneity. A non-plaque-purified baculovirus stock of recombinant chicken CMP-sialic acid: N-acetylgalactosaminide 2,6-sialyltransferase (ST6GalNAc-I) accession No. X74946
[GenBank]
was a gift from Dr Jim Paulson (Scripps Institute, La Jolla, CA). Viral stock was plaque-purified, amplified, and expressed in Sf9 cells. Enzyme was loaded onto an SP Sepharose FF cation-exchange column (Pharmacia, Piscataway, NJ) and washed with 5 column volume (CV) of buffer A (200 mM NaCl, 25 mM 2-[4-morpholino]-ethane sulfonic acid [MES], pH 6.0) to A280 baseline. The enzyme was eluted with a 20 CV elution gradient from 0 to 100% buffer B (1 M NaCl, 25 mM MES, pH 6.0). The active ST6GalNAc-I fraction pool from the SP Sepharose FF elution was adjusted to 0.5 M ammonium sulfate and loaded onto a Phenyl 650C column equilibrated with 5 CV of buffer A (0.75 M ammonium sulfate, 20 mM sodium phosphate, pH 7.0). The enzyme was loaded in 10 CV and washed to A280 baseline with 5 CV buffer A. The enzyme was eluted with a 10 CV elution gradient from 0 to 100% buffer B (20 mM sodium phosphate, pH 7.0). Active fractions from the Phenyl 650C step were pooled and concentrated to 4 mL with a 10K MWCO Amicon Ultra and loaded onto a Superdex 200 Hiload 16/60 column that was equilibrated with 1.5 CV of 150 mM NaCl, 50 mM Tris, pH 7.5 [Tris-buffered saline (TBS)]. The enzyme was eluted with 1.5 CV of TBS. The acceptor specificity of recombinant chicken CMP-NeuAc: GalNAc 2,6-sialyltransferase (ST6GalNAc-I) and its use for semi-preparative synthesis of glycopeptides have been described (Kurosawa et al., 1994; Blixt et al., 2002).
Core1 GalT-1. Drosophila UDP-Gal: GalNAc ß1,3 galactosyltransferase (Core1 GalT) (Ju et al., 2002) accession No. LD20186, truncated at E48, was subcloned in frame into pACGP67. This construct was transfected into Sf9 cells using the BaculoGold system (Invitrogen, Carlsbad, CA) and expressed in Sf9 cells according to the manufacturer’s instructions. Secreted enzyme was loaded onto a Q Sepharose FF column, which was washed with buffer A (25 mM NaCl, 25 mM Tris, pH 7.0, plus 0.01% Tween-80) until material absorbing at 280 nm was no longer detectable. The enzyme was eluted by a 20 CV gradient from 99% buffer A/1% buffer B (1 M NaCl, 25 mM Tris, pH 7.0 plus 0.01% Tween-80) to 100% buffer B. The active fractions were pooled, concentrated, and purified further using Superdex 200, as described above.
Screening of GalNAc-transferase isoforms for site-directed O-glycosylation
Semi-purified recombinant human GalNAc-Ts were tested for the capacity to transfer GalNAc to synthetic peptides derived from sequences covering potential O-glycan acceptor sites in proteins of interest. Screening enzyme reactions were performed with 0.1–0.5 mU GalNAc-Ts in 25 µL reaction mixtures containing 0.25% Triton X-100, 10 mM MnCl2, 100 µM UDP-[14C]-GalNAc (3200 cpm/nmol) (Amersham Pharmacia Biotech, Piscataway, NJ), 25 mM Na-cacodylate, pH 7.4, and 250 µM peptide substrate (Wandall et al., 1997). Reaction products were quantified by scintillation counting after chromatography on Dowex 1-X8 or analyzed by high-performance liquid chromatography (HPLC) and MALDI-TOF.
Glycosylation and PEGylation of E. coli-expressed cytokines
G-CSF
Recombinant G-CSF (960 µg) produced in E. coli was GalNAc glycosylated with GalNAc-T2 (40 mU) in a reaction mixture of 1 mL containing 9 mM UDP-GalNAc, 4 mM MnCl2, 25 mM MES buffer, pH 6.2, and 0.005% NaN3 at room temperature. MALDI-TOF/MS analysis showed the reaction was complete at 24 h. The product, (GalNAc)G-CSF, was purified by HPLC size exclusion chromatography on successive 10 x 300 mm Superdex 75 and Superdex 200 columns (Amersham) eluted at 1 mL/min with 0.15 M NaCl, 0.005% polysorbate 80, 20 mM sodium phosphate, pH 4.5, followed by concentration to a final volume of 1 mL using a Centricon filter (MWCO 5 kDa). The yield based on starting protein measured by absorbance at 280 nm was >95%. After buffer exchange into 25 mM MES, pH 6.2, and 0.005% NaN3, 1 mg of G-CSF(GalNAc) was incubated in a final volume of 1 mL containing 200 mU ST6GalNAc-I, 0.25 mM CMP-SiaPEG-20K, and 100 mM MnCl2, with gentle rocking at 32°C for 72 h. After concentration and buffer exchange into 25 mM NaOAc, pH 4.5, the product, G-CSF(GalNAc-SiaPEG-20K), was purified by HPLC on SP Sepharose eluted with 25 mM NaOAc, pH 4.5, followed by Superdex 75 eluted with 20 mM sodium phosphate, pH 7.2, 150 mM NaCl, and 0.005% polysorbate 80. The pooled product was concentrated as before and stored in the elution buffer at 4°C.
IFN-2b
Recombinant IFN-2b (4 mg) produced in E. coli was GalNAc-O-glycosylated with GalNAc-T2 (80 mU), 3 mM UDP-GalNAc, 150 mM NaCl, 5 mM MgCl2, 5 mM MnCl2, 0.05% polysorbate 80, 20 mM MES, pH 6.2, and 0.05% NaN3 in 2.12 mL at 32°C with slow rotary mixing. MALDI-TOF/MS analysis showed the reaction to be complete after 96 h. A 1-mL aliquot of the reaction mixture was buffer exchanged into 150 mM NaCl, 5 mM MgCl2, 5 mM MnCl2, 0.05% polysorbate 80, 20 mM MES (pH 7.4), and 0.05% NaN3 using a Centricon 5 kDa MWCO filter spin cartridge. The product, IFN-2b(GalNAc), was reconstituted into 1.2 mL of the same buffer containing 400 mU ST6GalNAc-I and 0.42 mM CMP-SiaPEG-20K, and the mixture was incubated at 32°C with slow rotary mixing for 96 h. The resulting product, IFN-2b(GalNAc-SiaPEG-20K), was purified by HPLC on SP Sepharose (HiTrap SP, FF 1 mL; Amersham) eluted with 25 mM sodium acetate plus 0.005% polysorbate 80 at 1 mL/min for 10 min, followed by gradient elution with 0–0.5 M NaCl in the same buffer. Pooled fractions were concentrated with 5 kDa MWCO spin filter, and aliquots (0.2 mg/200 µL) were further purified by HPLC on a Superdex 75 column (HR 10/30, 10 x 300 mm; Amersham) eluted with 150 mM NaCl, 0.005% polysorbate 80, 20 mM sodium phosphate, pH 6.2. Pooled fractions were concentrated as before and stored in the elution buffer at 4°C.
GM-CSF
Recombinant GM-CSF produced in E. coli (1 mg, 0.069 µmol) was incubated in a 1.3 mL aqueous solution containing 20 mM MES buffer, pH 6.0 plus 0.005% NaN3, 1.2 mM UDP-GalNAc, 80 mU GalNAc-T2, and 6.25 mM MnCl2 at room temperature. MALDI-TOF/MS indicated that the formation of GM-CSF(GalNAc)2 was complete, as measured by a molecular weight increase of 406 KDa, after 72 h. GM-CSF(GalNAc)2 (1 mg) was reconstituted into 1.25 mL 25 mM MES buffer containing 6 mg (9.8 mmol) UDP-Gal, 40 mU Core1 GalT-1, 6 mg (0.3 µmol) CMP-SiaPEG-20K, 120 mU ST3Gal-1, and 4 mM MnCl2. The resulting mixture was slowly rotated at room temperature for 24 h. Additional CMP-SiaPEG-20K (6 mg, 0.3 µmol) was added, and incubation continued at room temperature for an additional 24 h. The reaction mixture was concentrated to 1 mL, buffer exchanged with buffer A (25 mM NaOAc, 0.005% polysorbate 80, pH 6.0), and concentrated to 2.5 mL in a 5000 MWCO centrifugal filter. The resulting solution was purified on an Amersham HiTrap SP Sepharose FF column (5 mL) with isocratic elution of 100% buffer A for 10 min, followed by a linear gradient to 100% buffer B (2 M NaCl, 0.005% polysorbate 80, 25 mM NaOAc, pH 6.0) over 20 min at a flow rate of 3 mL/min. The peak eluting at 25% buffer B was collected and concentrated to 1.5 mL in a 5000 MWCO centrifugal filter. The concentrated protein solution was further purified on an Amersham HiLoad Superdex 200 (10 x 300 mm) eluted with 0.15 M NaCl plus 20 mM sodium phosphate, pH 5.0, and 0.005% polysorbate 80, at a flow rate of 0.3 mL/min. A peak containing the desired product eluting at 28.5 min was pooled and concentrated as before.
Structural analysis of peptides and proteins
MALDI-TOF mass spectrometry was performed on a Voyager-DE MALDI time-of-flight mass spectrometer (PerSeptive Biosystems, Foster City, CA) equipped with delayed extraction. Matrix for peptides was 2,5-dihydroxybenzoic acid, 25 g/L (Sigma, St. Louis, MO), dissolved in 30% aqueous acetonitrile. The MALDI matrix for the proteins was 3,5-dimethoxy-4-hydroxycinnamic acid (Sigma) 100 g/L dissolved in 30% acetonitrile in 0.3% trifluoroacetic acid. All mass spectra were obtained in the positive mode. Data processing was carried out using GRAMS/386 software.
For peptide mapping, protein samples were digested by GluC with or without trypsin overnight at 37°C and loaded on an LC-MS system (Finnigan LCQ-classic ion trap) with an electrospray ion source interfaced to a 15 cm x 300 µm id LC Packings PepMap reversed-phase capillary chromatography column. One microliter volume of the digest was injected, and the peptides were eluted from the column by an ACN/0.1% formic acid gradient at a flow rate of 3 µL/min. The electrospray ion source was operated at 4.0 kV.
G-CSF competition binding assay
A competition binding assay to estimate the affinity of G-CSF GlycoPEGylated variants to human G-CSF receptor was constructed using a truncated G-CSF receptor/IgG-Fc chimera immobilized on microtiter wells coated with protein A. Details of the assay are given in Supplementary Data.
G-CSF-dependent NFS-60 cellular proliferation assay
NFS-60 cells are responsive to interleukin-3, G-CSF, and macrophage colony stimulating factor (M-CSF) (Nakoinz et al., 1990) and are used in a cell proliferation assay described in Supplementary Data.
Viral inhibition assay
Viral inhibition titers for IFN-2b and derivatives were performed at PBL BioMedical Laboratory (Piscataway, NJ) using MDBK cells challenged with VSV according to procedures described by Familletti and others (1981). Samples were tested in duplicate against an international reference standard (human IFN-2b, NIH reference Gxa01-901-535).
Pharmacokinetic studies
Proteins were formulated in physiological saline, pH 6.5 plus 2.5% mannitol and 0.05% polysorbate 80. Samples containing [125I]-radiolabeled IFN-2b or IFN-2b(GalNAc-SiaPEG-20K) (10 µCi, 20 µCi/µg) (Amersham) or unlabeled G-CSF, G-CSF(GalNAc-SiaPEG-20K), or G-CSF(PEG-20K) (3 µg protein/rat) were injected as a bolus via the tail vein in rats (five animals per time point). Concentrations of injected proteins were measured by HPLC by comparing eluted sample peaks monitored at A280 nm with a bovine serum albumin (BSA) standard. At timed intervals after injection, whole blood samples were withdrawn via the jugular vein and assayed for radioactivity in a gamma counter to determine IFN-2b or IFN-2b(GalNAc-SiaPEG-20K), or heparinized plasma was assayed by enzyme-linked immunosorbent assay (ELISA) to determine concentrations of G-CSF, G-CSF(GalNAc-SiaPEG-20K), and G-CSF(PEG-20K). The ELISA (LOD 0.4 ng/mL) was performed utilizing goat anti-human G-CSF (R&D Catalog No. AF-214-NA) as a trapping antibody and europium-labeled murine anti-human G-CSF (R&D Catalog No. MAB214) as a detection antibody. The antibody was labeled with ITC-DTPA-Europium according to the manufacturer’s instructions (Perkin Elmer). All animal studies were conducted in compliance with the USDA Animal Welfare Act and the PHS Policy on Humane Care and Use of Laboratory Animals, and study protocols were reviewed by the Institutional Animal Care and Use Committees at Calvert Preclinical Services (Olyphant, PA).
Pharmacokinetic calculations
Mean serum radioactivity or drug concentration–time data for each compound were analyzed using noncompartmental methods (Gibaldi and Perrier, 1982). The terminal elimination rate constant (Ke) was estimated by log-linear regression of plasma concentration values in the apparent terminal elimination phase. The terminal elimination half-life (t1/2) was calculated as 0.693/Ke. Concentration values at time zero (C0) for area under the concentration–time curve (AUC) estimates were obtained by fitting concentration–time profiles to a one-compartment model and using the y-intercept for the time zero values. The AUC was estimated with the linear trapezoidal rule and extrapolated to time infinity (last concentration value/Ke). Clearance (Cl) was estimated by dividing dose by AUC0–
.
|
Supplementary Data |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and MethodsSupplementary DataConflict of interest statementAcknowledgmentsReferences |
|---|
Supplementary data are available at Glycobiology online (http://glycob.oxfordjournals.org/).
|
Conflict of interest statement |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and MethodsSupplementary DataConflict of interest statementAcknowledgmentsReferences |
|---|
None declared.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and MethodsSupplementary DataConflict of interest statementAcknowledgmentsReferences |
|---|
We thank Bruce Mico and Tom Stevenson for helpful insights regarding preparation of this manuscript and GloriaMay Machado for excellent technical assistance. Funding for this work was provided to Henrik Clausen by the Danish Cancer Society and the Danish Medical Research Council. Funding to pay the Open Access publication charges for this article was provided by Neose Technologies, Inc.
|
Abbreviations |
|---|
AUC, area under the concentration–time curve; CMP-SiaPEG, cytidine monophosphate 5'-aminoglycyl-methoxypolyeneglycol neuraminic acid; GalNAc-T, UDP-GalNAc: polypeptide
-N-acetylgalactosaminyltransferase; G-CSF, granulocyte colony stimulating factor; GM-CSF, granulocyte/macrophage colony stimulating factor; HPLC, high-performance liquid chromatography; IFN-2b, interferon-alpha2b; IgG, immunoglobulin G; LC-MS/MS, liquid chromatography-mass spectrometry/mass spectrometry; MALDI-TOF/MS, matrix-assisted laser desorption ionization-time-of-flight mass spectrometry; MDBK, Madin–Darby bovine kidney; MES, 2-(4-morpholino)-ethane sulfonic acid; PEG, polyethylene glycol; SDS–PAGE, sodium dodecyl sulfate–polyacrylamide gel electrophoresis; VSV, vesicular stomatitis virus; WBC, white blood cell. |
References |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and MethodsSupplementary DataConflict of interest statementAcknowledgmentsReferences |
|---|
Aritomi, M., Kunishima, N., Okamoto, T., Kuroki, R., Ota, Y., and Morikawa, K. (1999) Atomic structure of the GCSF-receptor complex showing a new cytokine-receptor recognition scheme. Nature, 401, 713–717.[CrossRef][Medline]
Bennett, E.P., Hassan, H., Mandel, U., Hollingsworth, M.A., Akisawa, N., Ikematsu, Y., Merkx, G., van Kessel, A.G., Olofsson, S., and Clausen, H. (1999) Cloning and characterization of a close homologue of human UDP-N-acetyl-alpha-D-galactosamine: polypeptide N-acetylgalactosaminyltransferase-T3, designated GalNAc-T6. Evidence for genetic but not functional redundancy. J. Biol. Chem., 274, 25362–25370.
Bennett, E.P., Hassan, H., Mandel, U., Mirgorodskaya, E., Roepstorff, P., Burchell, J., Taylor-Papadimitriou, J., Hollingsworth, M.A., Merkx, G., van Kessel, A.G., and others (1998) Cloning of a human UDP-N-acetyl-alpha-D-galactosamine: polypeptide N-acetylgalactosaminyltransferase that complements other GalNAc-transferases in complete O-glycosylation of the MUC1 tandem repeat. J. Biol. Chem., 273, 30472–30481.
Blixt, O., Allin, K., Pereira, L., Datta, A., and Paulson, J.C. (2002) Efficient chemoenzymatic synthesis of O-linked sialyl oligosaccharides. J. Am. Chem. Soc., 124, 5739–5746.[CrossRef][ISI][Medline]
Bodo, G. and Maurer-Fogy, I. (1985a) Molecular Species of recombinant human Interferon Alpha 2 detected in affinity purified preparations. In Stewart, W. and Shellekens, H. (eds), The Biology of the Interferon System. Elsevier, Amsterdam, pp. 59–64.
Bodo, G. and Maurer-Fogy, I. (1985b) Characterization of different molecular species in affinity purified recombinant human Interferon Alpha 2. The Interferon System. Serono Symposium Publication, Raven Press, New York, pp. 23–27.
DeFrees, S., Zopf, D., Bayer, R., Bowe, C., Hakes, D., and Chen, X. (2004) Glycopegylation methods and proteins/peptides produced by the methods. US patent application 20040132640. 2004 Aug 7.
Familletti, P.C., Rubinstein, S., and Pestka, S. (1981) A convenient and rapid cytopathic effect inhibition assay for interferon. In Pestka, S. (ed.), Methods in Enzymology. Academic Press, New York, pp. 387–394.
Forno, G., Fogolin, M.B., Oggero, M., Kratje, R., Etcheverrigaray, M., Conradt, H.S., and Nimtz, M. (2004) N- and O-linked carbohydrates and glycosylation site occupancy in recombinant human granulocyte-macrophage colony-stimulating factor secreted by a Chinese hamster ovary cell line. Eur. J. Biochem., 271, 907–919.[ISI][Medline]
Foser, S., Schacher, A., Weyer, K.A., Brugger, D., Dietel, E., Marti, S., and Schreitmuller, T. (2003) Isolation, structural characterization, and antiviral activity of positional isomers of monopegylated interferon alpha-2a (PEGASYS). Protein Expr. Purif., 30, 78–87.[CrossRef][ISI][Medline]
Francis, G.E., Fisher, D., Delgado, C., Malik, F., Gardiner, A., and Neale, D. (1998) PEGylation of cytokines and other therapeutic proteins and peptides: the importance of biological optimisation of coupling techniques. Int. J. Hematol., 68, 1–18.[CrossRef][ISI][Medline]
Gibaldi, M. and Perrier, D. (1982) Pharmacokinetics. Marcel Dekker, New York.
Gross, H.J., Rose, U., Krause, J.M., Paulson, J.C., Schmid, K., Feeney, R.E., and Brossmer, R. (1989) Transfer of synthetic sialic acid analogues to N- and O-linked glycoprotein glycans using four different mammalian sialyltransferases. Biochemistry, 28, 7386–7392.[CrossRef][Medline]
Harris, J.M. and Chess, R.B. (2003) Effect of pegylation on pharmaceuticals. Nat. Rev. Drug Discov., 2, 214–221.[CrossRef][ISI][Medline]
Harris, J.M., Martin, N.E., and Modi, M. (2001) Pegylation: a novel process for modifying pharmacokinetics. Clin. Pharmacokinet., 40, 539–551.[CrossRef][ISI][Medline]
Hassan, H., Bennett, E.P., Mandel, U., Hollingsworth, M.A., and Clausen, H. (2000) Carbohydrates in Chemistry and Biology – A Comprehension Handbook. Wiley-VCH, New York, pp. 273–292.
Jeanneau, C., Chazalet, V., Auge, C., Soumpasis, D.M., Harduin-Lepers, A., Delannoy, P., Imberty, A., and Breton, C. (2004) Structure-function analysis of the human sialyltransferase ST3Gal I: role of N-glycosylation and a novel conserved sialylmotif. J. Biol. Chem., 279, 13461–13468.
Jolling, K., Ruixo, J.J., Hemeryck, A., Piotrovskij, V., and Greway, T. (2004) Population pharmacokinetic analysis of pegylated human erythropoietin in rats. J. Pharm. Sci., 93, 3027–3038.[CrossRef][ISI][Medline]
Ju, T., Brewer, K., D’Souza, A., Cummings, R.D., and Canfield, W.M. (2002) Cloning and expression of human core 1 beta1,3-galactosyltransferase. J. Biol. Chem., 277, 178–186.
Katre, N.V. (1993) The conjugation of proteins with polyethylene glycol and other polymers: altering properties of proteins to enhance their therapeutic potential. Adv. Drug Deliv. Rev., 10, 91–114.[CrossRef]
Kaushansky, K., Lopez, J.A., and Brown, C.B. (1992) Role of carbohydrate modification in the production and secretion of human granulocyte macrophage colony-stimulating factor in genetically engineered and normal mesenchymal cells. Biochemistry, 31, 1881–1886.[CrossRef][Medline]
Kaushansky, K., O’Hara, P.J., Hart, C.E., Forstrom, J.W., and Hagen, F.S. (1987) Role of carbohydrate in the function of human granulocyte-macrophage colony-stimulating factor. Biochemistry, 26, 4861–4867.[CrossRef][Medline]
Kinstler, O.B., Brems, D.N., Lauren, S.L., Paige, A.G., Hamburger, J.B., and Treuheit, M.G. (1996) Characterization and stability of N-terminally PEGylated rhG-CSF. Pharm. Res., 13, 996–1002.[CrossRef][ISI][Medline]
Kurosawa, N., Hamamoto, T., Lee, Y.C., Nakaoka, T., Kojima, N., and Tsuji, S. (1994) Molecular cloning and expression of GalNAc alpha 2,6-sialyltransferase. J. Biol. Chem., 269, 1402–1409.
Kurosawa, N., Takashima, S., Kono, M., Ikehara, Y., Inoue, M., Tachida, Y., Narimatsu, H., and Tsuji, S. (2000) Molecular cloning and genomic analysis of mouse GalNAc alpha 2,6-sialyltransferase (ST6GalNAc I). J. Biochem., 127, 845–854.
Kuwabara, T., Uchimura, T., Kobayashi, H., Kobayashi, S., and Sugiyama, Y. (1995) Receptor-mediated clearance of G-CSF derivative nartograstim in bone marrow of rats. Am. J. Physiol., 269, E1–E9.
Lord, B.I., Woolford, L.B., and Molineux, G. (2001) Kinetics of neutrophil production in normal and neutropenic animals during the response to filgrastim (r-metHu G-CSF) or filgrastim SD/01 (PEG-r-metHu G-CSF). Clin. Cancer Res., 7, 2085–2090.
Nakoinz, I., Lee, M.T., Weaver, J.F., and Ralph, P. (1990) Differentiation of the IL-3-dependent NFS-60 cell line and adaption to growth in macrophage colony-stimulating factor. J. Immunol., 145, 860–864.[Abstract]
Neupogen (2005) Neupogen Prescribing Information, available at http://www.neupogen.com/pi.html/.
Radhakrishnan, R., Walter, L.J., Hruza, A., Reichert, P., Trotta, P.P., Nagabhushan, T.L., and Walter, M.R. (1996) Zinc mediated dimer of human interferon-alpha 2b revealed by X-ray crystallography. Structure, 4, 1453–1463.[Medline]
Roberts, M.J., Bentley, M.D., and Harris, J.M. (2002) Chemistry for peptide and protein PEGylation. Adv. Drug Deliv. Rev., 54, 459–476.[CrossRef][ISI][Medline]
Rotondaro, L., De Paolis, E., Ferrero, D., D’Alatri, L., Raucci, G., Fabbri, A., Gerwig, G.J., Kamerling, J.P., Mariani, M.F., Mele, A., and others (1999) Purification and characterization of two recombinant human granulocyte colony-stimulating factor glycoforms. Pharmacokinetic and activity studies of single-dose administration in mice. Mol. Biotechnol., 11, 117–128.[CrossRef][ISI][Medline]
Sato, H. (2002) Enzymatic procedure for site-specific pegylation of proteins. Adv. Drug Deliv. Rev., 54, 487–504.[CrossRef][ISI][Medline]
Schwientek, T., Bennett, E.P., Flores, C., Thacker, J., Hollmann, M., Reis, C.A., Behrens, J., Mandel, U., Keck, B., Schafer, M.A., and others (2002) Functional conservation of subfamilies of putative UDP-N-acetylgalactosamine: polypeptide N-acetylgalactosaminyltransferases in Drosophila, Caenorhabditis elegans, and mammals. One subfamily composed of l(2)35Aa is essential in Drosophila. J.Biol. Chem., 277, 22623–22638.
Tanaka, H. and Tokiwa, T. (1990) Pharmacokinetics of recombinant human granulocyte colony-stimulating factor studied in the rat by a sandwich enzyme-linked immunosorbent assay. J. Pharmacol. Exp. Ther., 255, 724–729.
Ten Hagen, K.G., Fritz, T.A., and Tabak, L.A. (2003) All in the family: the UDP-GalNAc: polypeptide N-acetylgalactosaminyltransferases. Glycobiology, 13, 1R–16R.
Thomas, L.J., Panneerselvam, K., Beattie, D.T., Picard, M.D., Xu, B., Rittershaus, C.W., Marsh, H.C., Jr., Hammond, R.A., Qian, J., Stevenson, T., and others (2004) Production of a complement inhibitor possessing sialyl Lewis X moieties by in vitro glycosylation technology. Glycobiology, 14, 883–893.
Tsuboi, S., Srivastava, O.P., Palcic, M.M., Hindsgaul, O., and Fukuda, M. (2000) Acquisition of P-selectin binding activity by en bloc transfer of sulfo Le(x) trisaccharide to the cell surface: comparison to a sialyl Le(x) tetrasaccharide transferred on the cell surface. Arch. Biochem. Biophys., 374, 100–106.[CrossRef][ISI][Medline]
Wandall, H.H., Hassan, H., Mirgorodskaya, E., Kristensen, A.K., Roepstorff, P., Bennett, E.P., Nielsen, P.A., Hollingsworth, M.A., Burchell, J., Taylor-Papadimitriou, J., and others (1997) Substrate specificities of three members of the human UDP-N-acetyl-alpha-D-galactosamine: Polypeptide N-acetylgalactosaminyltransferase family, GalNAc-T1, -T2, and -T3. J. Biol. Chem., 272, 23503–23514.
Warnock, D., Bai, X., Autote, K., Gonzales, J., Kinealy, K., Yan, B., Qian, J., Stevenson, T., Zopf, D., and Bayer, R.J. (2005) In vitro galactosylation of human IgG at 1 kg scale using recombinant galactosyltransferase. Biotechnol. Bioeng., 92, 831–842.
Yang, K., Basu, A., Wang, M., Chintala, R., Hsieh, M.C., Liu, S., Hua, J., Zhang, Z., Zhou, J., Li, M., and others (2003) Tailoring structure-function and pharmacokinetic properties of single-chain Fv proteins by site-specific PEGylation. Protein Eng., 16, 761–770.
Yoshioka, Y., Tsutsumi, Y., Ikemizu, S., Yamamoto, Y., Shibata, H., Nishibata, T., Mukai, Y., Okamoto, T., Taniai, M., Kawamura, M., and others (2004) Optimal site-specific PEGylation of mutant TNF-alpha improves its antitumor potency. Biochem. Biophys. Res. Commun., 315, 808–814.[CrossRef][ISI][Medline]
Youngster, S., Wang, Y.S., Grace, M., Bausch, J., Bordens, R., and Wyss, D.F. (2002) Structure, biology, and therapeutic implications of pegylated interferon alpha-2b. Curr. Pharm. Des., 8, 2139–2157.[CrossRef][ISI][Medline]
Zamboni, W.C. (2003) Pharmacokinetics of pegfilgrastim. Pharmacotherapy, 23, 9S–14S.[CrossRef][ISI][Medline]