ARTICLE pubs.acs.org/bc
Aldehyde PEGylation Kinetics: A Standard Protein versus a Pharmaceutically Relevant Single Chain Variable Fragment Anna Moosmann,*,#,† Jessica Blath,#,‡ Robert Lindner,† Egbert M€uller,§ and Heiner B€ottinger† †
Institute of Cell Biology and Immunology, University of Stuttgart, Allmandring 31, 70569 Stuttgart, Germany Institute of Interfacial Engineering, University of Stuttgart, Nobelstrasse 12, 70569 Stuttgart, Germany § Tosoh Bioscience GmbH, Zettachring 6, 70567 Stuttgart, Germany ‡
ABSTRACT:
The mPEG-aldehyde PEGylation with two different PEG sizes and two proteins was experimentally determined with respect to yield, conversion, and selectivity. The kinetic behavior of these PEGylation reactions was simulated using a numerically solved set of differential equations. We show that the assumption of an inactivation of mPEG-aldehyde is crucial for the simulation of the overall PEGylation and that the inactivation is pH-dependent. We further demonstrate that ideal PEGylation parameters such as pH, temperature, reaction time, and protein concentration need to be chosen carefully depending on the protein and PEG size. In terms of selectivity and yield, we show that the reaction should be stopped before the highest mono-PEG concentration is reached. Moreover, room temperature and a slightly acidic pH of approximately 6 are good starting points. In conclusion, selectivity can be optimized choosing a shorter reaction time and a reduced reaction temperature.
’ INTRODUCTION There are numerous drugs based on proteins of different classes, such as enzymes and antibodies, both having great potential in pharmaceutical use. However, the usage of proteins in general is restricted by their biocompatibility and in situ halflife. In particular, small molecules have very short half-lives, which restricts the use of small proteins in drug design.1,2 For example, antibody fragments have many characteristics that would make them ideal drug candidates, such as easy and cheap production and high efficacy; however, due to their small size blood and renal clearance are very high.1,2 An effective way of prolonging the in situ half-life of such proteins is polymer modification, which artificially enlarges the hydrodynamic radius and leads to a masking effect.25 As the renal clearance depends highly on the hydrodynamic radius, renal filtration as a fast excretion step will be overcome. The masking effect by polymer modification will also decrease the target sites for proteases1,2,6 and thereby prolong the half-life, too. One possible way of polymer modification is PEGylation, which denominates the chemical attachment of a poly(ethylene glycol) chain to molecules such as proteins or peptides. Since the first reports in PEGylation in the 1970s,3,7 the method has now become a state-of-the-art technology. Eight PEGylated r 2011 American Chemical Society
protein drugs are approved by the FDA by now,8 such as PEGasys (Hoffman-La Roche) and PEG Intron (ScheringPlough/Enzon),9,10 both R-Interferon products for hepatitis C treatment. PEG is the polymer of choice because it is a neutral, nonimmunogenic, and inert polymer.6,11,12 PEG has to be functionalized with a reactive end group to enable attachment to another molecule. Each functionalization of PEG results in altered PEGylation kinetics, but additionally the protein itself affects the kinetic characteristics of the PEGylation reaction as well as the ambient conditions. One of the most frequently used functionalities is aldehyde-modified PEG, which reacts with free amino groups of the protein.6,8 This reaction is random, as all available amino groups can react. The pKa and the accessibility of each free amino group define the probability for PEGylation.2,6,8,13,14 These characteristics can be influenced by the ambient conditions. The pH is a critical factor in PEGylation characteristics, as well as temperature and concentrations of protein and PEG.6,8,15 PEGylation changes the physicochemical properties of a protein,9,16 and the Received: February 21, 2011 Revised: July 18, 2011 Published: July 24, 2011 1545
dx.doi.org/10.1021/bc200090x | Bioconjugate Chem. 2011, 22, 1545–1558
Bioconjugate Chemistry chromatographic behavior of proteins is modified by PEGylation, too. The altered behavior of PEGylated proteins in comparison with the non-PEGylated parental proteins could be shown by ion-exchange chromatography (IEX),9,16,17 hydrophobic interaction chromatography,9 as well as size exclusion chromatography (SEC).9 In this study, we thus focus on IEX for the rapid separation of PEGylated and non-PEGylated proteins to follow the PEGylation reaction as fast as possible. Furthermore, studies on the standard protein lysozyme have already been performed regarding its behavior in chromatography,18,19 as well as accessible PEGylation sites for aldehyde chemistry,20 but thus far, no detailed survey on the kinetics of the aldehyde PEGylation reaction itself has been published. We compare the PEGylation kinetics of two different proteins, a pharmaceutically relevant single-chain variable fragment produced in E. coli and the enzyme lysozyme. Both proteins were PEGylated and the reaction was closely observed to obtain the relevant data for kinetic analysis. Different factors influencing the environmental conditions such as pH and temperature have additionally been analyzed and two different PEG sizes were used. Finally, to gather deeper insight into the reaction kinetics, a model was developed to simulate the reaction behavior.
’ MATERIAL AND METHODS Chemicals. Methoxy-PEG-aldehyde (mPEG-AL) with average molecular weights of 5 and 30 kDa were purchased from NOF Corp. (Grobbendonk, Belgium). Lysozyme (98% pure, chicken egg white) was provided by Sigma (St. Louis, USA). All other chemicals were provided by Merck (Darmstadt, Germany). Production of scFv. The single-chain variable fragment used in this study was produced in E. coli BL21 DE3 rha- using the rhamnose induction system.21,22 Following a modified osmotic shock procedure according to Rathore,23 the pellet was solubilized in 100 mM phosphate buffer containing 50% glucose followed by adding pure water. After centrifugation, the protein solution was purified using protein L-chromatography (GenScript, Piscataway, USA). The purified scFv was further diafiltered and lyophilized for long-term storage. PEGylation of scFv and Lysozyme. Protein and 5 or 30 kDa PEG of the desired concentration were dissolved in a 20 mM sodium phosphate buffer pH 6.0 or 7.0 or a 20 mM sodium acetate buffer pH 3.0, 4.0, or 5.0 containing 20 mM NaCNBH3.4,18,19 The PEGylation reaction was performed in a final volume of 1 mL at temperatures between 10 and 30 °C and a protein concentration in the range 110 g/L. The reaction time was up to 20 h; the monitoring was carried out continuously. All discussed factors were analyzed individually; each reaction was replicated at least two times. The following standard conditions were used as starting conditions for each factor analysis: for lysozyme, pH 6, 21 °C, 20 h reaction time, 2-fold PEG excess, 1 g/L; for scFv, 1 g/L, pH 4, 21 °C, 20 h reaction time, 5-fold PEG excess. Lysozyme and scFv both possess a N-terminal amino group, and lysozyme possesses 6 lysine residues whereas the scFv has 9. Therefore, more than one mono-PEGylated isoform and also different di-PEGylated forms as well as higher PEGylated species can be formed in the random reaction. Analytical Procedures. An analytical TSKgel SP-NPR column (ID 4.6 mm L 35 mm, Tosoh Bioscience GmbH, Stuttgart, Germany) was used to track the PEGylation reaction. For lysozyme, the buffer consisted of 25 mM sodium phosphate buffer, pH 6.0. For elution, 1 M NaCl was added. For scFv, the
ARTICLE
buffer consisted of 20 mM sodium acetate buffer, pH 4.5. For elution, 1 M NaCl was added. The analytical IEX was carried out on a Thermo Separation HPLC SpectraSYSTEM (Thermo Fisher Scientific GmbH, Dreieich, Germany). SDS-PAGE under reducing conditions was performed according to Laemmli.24 Protein samples were solubilized in sample buffer and heated at 95 °C for 3 min. SDS-PAGE was performed with precast NuPAGE Novex 10% Bis-Tris Midigels (Invitrogen Corporation, Carlsbad, USA) in an XCell4 SureLock Midi-Cell (Invitrogen) according to the manufacturer’s procedure. The gels were stained with PageBlue Protein Staining solution (Fermentas, St. Leon-Rot, Germany) according to the manufacturer’s instruction. To stain specifically the PEGylated proteins, bariumiodine staining was carried out. The procedure was modified according to Kurf€urst,25 following the instructions of Bailon et al.26 SEC was carried out on an analytical TSKgel G3000SWXL column (7.8 mm 30 cm, Tosoh Bioscience GmbH). As mobile phase, a 100 mM sodium phosphate buffer, pH 6.7, containing 100 mM Na2SO4 and 0.05% NaN3 was used. The SEC chromatography was performed on a Thermo Separation HPLC SpectraSYSTEM (Thermo Fisher Scientific GmbH). Mathematic Modeling. For the PEGylated forms and the unmodified forms at all time points, the concentration was estimated via the peak area from IEX chromatography, under the assumption of PEG being undetectable for UV absorbance. As PEG is undetectable, no experimental data for the PEG concentration were utilizable; rather, the PEG concentration, needed for the modeling, was calculated from the protein concentration by the following equation: cðPEGÞt ¼ cðPEGÞ0 cðmono-PEG-proteinÞt 2 cðdi-PEG-proteinÞt 3 cðtri-PEG-proteinÞt The set of differential equations describing the reaction (see Kinetic Studies and Mathematic Modeling section) was solved numerically with the program MatLab (v R2010a, MathWorks, Ismaning, Germany).14 The MatLab ode23s solver implemented in our case uses a modified Rosenbrock formula of order 2. Furthermore, an optimization was carried out using the MatLab toolbox fminsearch to fit the reaction kinetic parameters to the measured data. Calculation of Conversion, Yield, and Selectivity. The calculation of conversion, yield, and selectivity followed the conventions.27 Conversion X: X ¼
cE 0 cE cE 0
where cE = reactant concentration and c0E = reactant concentration at start of reaction. Yield Y: cP Y ¼ 0 cE where cP = product concentration (Mono-PEGylated protein for the present study). Selectivity S: S¼ 1546
Y X dx.doi.org/10.1021/bc200090x |Bioconjugate Chem. 2011, 22, 1545–1558
Bioconjugate Chemistry
ARTICLE
Figure 1. PEGylation reaction of lysozyme and 5 kDa PEG after 0, 80, and 350 min reaction time. Analyzed via cation-exchange and SDS-PAGE. Peaks and bands were labeled as results of MALDI-TOF and SDS-PAGE indicated. Left chromatogram, reaction time 0 min; center chromatogram, reaction time 80 min; right chromatogram, reaction time 350 min. Blue, Coomassie stained proteins; green, PEGylated proteins Coomassie and bariumiodine stained. For chromatographic conditions, see Analytical procedures.
Figure 2. PEGylation reaction of lysozyme with 5 and 30 kDa PEG after 20 h reaction time. Analyzed via cation-exchanger. Peaks were labeled as results of MALDI-TOF and SDS-PAGE indicated. Black line, 5 kDa PEGylated lysozyme; gray line, 30 kDa PEGylated lysozyme. For chromatographic conditions, see Analytical procedures.
’ RESULTS AND DISCUSSION PEGylation of Lysozyme. The PEGylation reaction was closely followed using an analytical SP-NPR column (Tosoh Bioscience GmbH) with a total analysis time of 6 min. Chromatograms were evaluated using the peak area as given by the program ChromQuest (Thermo Separations). Figure 1 shows typical chromatograms and the corresponding bands in SDSPAGE of lysozyme and PEGylated lysozyme before the start of the reaction and after a reaction time of 80 and 350 min with a 5 kDa PEG. Lysozyme remained in the mixture, as PEG did, which is not visible in UV and thereby not shown. In SDS-PAGE, PEG of 5 kDa was not visible because it migrated out of the gel. During the reaction, two isoforms of Mono-PEGylated lysozyme were formed, as well as a Di- and a Tri-PEGylated form.4,20 The Tri-PEGylated form did appear after more than 350 min and is not shown in Figure 1. The other PEGylation products are shown in Figure 1 to clarify the peak identification process. In a former study,18 fractions of the elution peaks shown in Figure 1 were collected and sent to the Friedrich-Alexander University (Erlangen, Germany) to perform the MALDI-TOF analysis.4 The peaks were identified as labeled in Figure 1 with the help of
SDS-PAGE and MALDI-TOF. Analogue experiments were done for PEGylation with 30 kDa PEG and the results were comparable, but the retention times were slightly shorter. A resolution comparison between a 5 and 30 kDa PEGylated lysozyme is shown in Figure 2. The resolution was not affected by the different sizes used in this work. In prior work, it could already be shown that the retention times in IEX decreased with increasing PEG length and increasing number of added PEG chains.16,18,29 Additionally, SDS-PAGE with the reaction mixtures was carried out after 20 h (Figure 3). The Coomassie staining revealed the protein fractions in blue, whereas a bariumiodine staining showed PEG in brown. For the PEGylated proteins, a mixture of both colors can be seen, resulting in green to brownish bands (Figures 1 and 3). It is well-known that PEG in SDS-PAGE migrates more slowly than proteins, so the protein standard used leads to higher molecular weights of PEG and PEGylated proteins.8,15,26 PEG of 5 kDa size is not visible in Figure 3 because it migrated out of the gel during the washing and staining procedure, but it was visible after the very first bariumiodine staining (results not shown). The apparent size of 5 kDa PEG in SDS-PAGE was estimated to be 8 kDa; for 30 kDa PEG, the apparent size in SDS-PAGE was about 50 kDa. Wiyh these values 1547
dx.doi.org/10.1021/bc200090x |Bioconjugate Chem. 2011, 22, 1545–1558
Bioconjugate Chemistry
ARTICLE
Figure 3. SDS-PAGE of a lysozyme PEGylation with 5 kDa PEG (left) and 30 kDa PEG (right) at different pH values after a reaction time of approximately 1200 min. Blue, Coomassie stained proteins; brown, bariumiodine stained PEG. Mixed colors represent PEGylated proteins. The marker proteins are on the left side.
Table 1. Relative Migration Length (Rf), Theoretical Size, and Calculated Size afor Analyzed Samples of PEGylated Lysozyme theoretical apparent size of PEGylated Rf [cm]
sample
theoretical size [kDa]
calculated size [kDa]
protein in SDS-PAGE [kDa] 214.4
0.55
Poly-PEG-30-Lys
134.4
588.45
0.9
Tri-PEG-30-Lys
104.4
258.07
164.4
1.35
Di-PEG-30-Lys
74.4
130.92
114.4
2
Mono-PEG-30-Lys
44.4
67.81
64.4
2.4
PEG 30 kDa
30
49.98
50
2.8
Tri-PEG-5-Lys
29.4
38.61
38.4
3.3 4
Di-PEG-5-Lys PEG-5-Lys
24.4 19.4
29.33 21.26
30.4 22.4
4.95
Lysozym
14.4
14.88
14.4
a
Column 4 calculated with the protein marker only; column 5 calculated theoretical apparent size taking into account the calculated size of PEG (column 4) as it appeared in SDS-PAGE.
taken into account, a new theoretical size in SDS-PAGE could be estimated (Table 1). This estimated size correlated well with the measured size gained with the help of the protein standard. The bigger the size of the PEGylated proteins, the more imprecise the size estimation was; however, this could be a gel-specific problem and was therefore not further investigated. The MALDI-TOF analyses confirmed the results of the SDS-PAGE. Thus, SDS-PAGE can alternatively be used as a tool for size estimation of PEGylated proteins and peak identification in IEX. PEGylation of scFv. The PEGylation reaction was followed as described for lysozyme. Again, SDS-PAGE at different time points correlated the peaks with the bands, both appearing after varying reaction times (Figure 5). The scFv monomer and dimer peak were identified using SEC because the dimerized scFv was not stable in the presence of SDS (data not shown). In the course of the PEGylation reaction, the dimer of the scFv disappeared and several peaks of shorter retention times appeared (Figure 5). This matches the chromatographic behavior of PEGylated lysozyme. Retention times in IEX decreased the higher the PEGylation degree got.16,18,29 Small peaks with a longer retention time than scFv also emerged during the PEGylation reaction. These peaks were not visible in a SDS-PAGE and thereby
could be aggregated proteins, PEGylated dimer-scFv, or small amounts of PEGylated impurities. The SDS-PAGE revealed clear bands, matching the presumption, as Mono-, Di-, and TriPEGylated forms were clearly visible and the sizes could be estimated as defined for lysozyme (see Table 2). For the PEGylation with 30 kDa PEG, an unexpected band appeared above the Mono-PEGylated scFv, which was light brown (Figure 4). The absence of blue color is evidence for the absence of a protein fraction. In addition, the calculated size of about 116 kDa indicated that this band was a PEG 30 kDa dimer, which should migrate at about 100 kDa in SDS-PAGE. This assumption was further supported by the slightly visible band in Figure 3 below the Di-PEG-30-lysozyme band, and the appearance of this band in SDS-PAGE of both proteins confirmed that it was not a protein-specific phenomenon. To verify the assumption of having a dimer PEG 30 kDa, another SDS-PAGE was carried out with PEG of 5 and 30 kDa in PEGylation reaction buffer of pH 5 and NaCNBH3 without the presence of protein (Figure 6). For 30 kDa PEG, a second band was clearly visible, for 5 kDa PEG the second band was barely visible and disappeared very fast. This is evidence for a dimerization of PEG under the present reaction conditions. 1548
dx.doi.org/10.1021/bc200090x |Bioconjugate Chem. 2011, 22, 1545–1558
Bioconjugate Chemistry
ARTICLE
Figure 4. SDS-PAGE of a scFv PEGylation with 5 kDa PEG (left) and 30 kDa PEG (right) at different protein-to-PEG ratios after a reaction time of approximately 1200 min. Blue, Coomassie stained proteins; brown, bariumiodine stained PEG. Mixed colors represent PEGylated proteins. The marker proteins are on the left side.
Figure 5. PEGylation reaction of scFv and 5 kDa PEG after 0, 80, and 350 min reaction time. Analyzed via cation-exchanger and SDS-PAGE. Peaks and bands were labeled as results of SDS-PAGE indicated. Left chromatogram, reaction time 0 min; center chromatogram, reaction time 80 min; right chromatogram, reaction time 350 min. Blue, Coomassie stained proteins; green, PEGylated proteins Coomassie and bariumiodine stained. For chromatographic conditions, see Analytical procedures.
dimer peaks of scFv were also summed up to unmodified protein. Peak areas were used to determine the concentrations of the PEGylated forms. In the case of the unmodified proteins, the concentration in the beginning of the reaction was known. In a first attempt, the following reaction model was stated:
Figure 6. SDS-PAGE of 30 (left) and 5 kDa PEG (right) solubilized in reaction buffer at pH 5. Black: bariumiodine stained PEG.
According to this result, the peaks of the IEX chromatogram were labeled as shown in Figure 4. For the PEGylated scFv, only one peak containing Mono-PEGylated protein was visible. The PEGylation of the scFv with 30 kDA PEG showed an analogue performance to lysozyme. Kinetic Studies and Mathematic Modeling. To monitor the PEGylation reaction all Mono-PEGylated isoforms were summed up to Mono-PEG-protein and the monomer and
PEG þ protein f Mono-PEG-protein
ð1Þ
Mono-PEG-protein þ PEG f Di-PEG-protein
ð2Þ
Di-PEG-protein þ PEG f Tri-PEG-protein
ð3Þ
The reaction rates r were thereby defined: r 1 ¼ PEG protein k1 r 2 ¼ PEG Mono-PEG-protein k2 r 3 ¼ PEG Di-PEG-protein k3 1549
dx.doi.org/10.1021/bc200090x |Bioconjugate Chem. 2011, 22, 1545–1558
Bioconjugate Chemistry
ARTICLE
Figure 7. PEGylation reaction for lysozyme and 5 kDa PEG at a pH of 4 (top left), at a pH of 7 (top right), scFv, and 5 kDa PEG at pH 4 (bottom left) and 7 (bottom right). For lysozyme PEGylation, 21 °C, 20 h reaction time, 2-fold PEG excess, 1 g/L protein; and for scFv, 1 g/L protein, 21 °C, 20 h reaction time, 5-fold PEG excess. Experimental data shown as crosses; simulated data shown as straight lines. Green, unmodified protein; red, MonoPEG-protein; blue, Di-PEG-protein; yellow, Tri-PEG-protein.
and the differential equations were specified: dxðPEGÞ ¼ r 1 r 2 r 3 dxðproteinÞ ¼ r 1 dxðMono-PEG-proteinÞ ¼ r 1 r 2 dxðDi-PEG-proteinÞ ¼ r 2 r 3 dxðTri-PEG-proteinÞ ¼ r 3 With the program MatLab, the differential equations were numerically solved, the experimental data was compared to the calculated results, and an optimization was carried out as specified.14 As a result, MatLab showed the experimental data as crosses and the calculated concentrations as lines (exemplarily shown in Figure 7 for four data sets). The first attempt at reaction modeling seemed to work well at a pH of 7 (Figure 7, right side). However, the more acidic the pH got, the worse the simulated reaction fitted the data (Figure 7 left column). The real reaction worked faster on a short time scale and then reached a plateau.
With the knowledge that dimerized PEG could be detected in SDS-PAGE (Figure 6), the next step was to include another reaction that inactivates PEG with the ability to form a PEG molecule that can no longer react. For example, a reaction with NaCNBH3 or an aldol reaction leading to PEG-multimers is a possibility for PEG inactivation. In SDS-PAGE, dimers were detected, but they appeared at different pH values (see Figure 6 for pH 5), whereas the kinetic simulation seems to favor a reaction that is pH dependent. Thus, an aldol reaction seems to be unlikely, but the nature of the inactivation reaction was not further observed. In a second attempt, an undefined inactivation step for PEG was included in the reaction set as done by Buckley and co-workers for mPEGsuccinimidyl propionate:14
1550
PEG þ protein f Mono-PEG-protein
ð1Þ
Mono-PEG-protein þ PEG f Di-PEG-protein
ð2Þ
Di-PEG-protein þ PEG f Tri-PEG-protein
ð3Þ
PEGreacting f PEGinactive
ð4Þ
dx.doi.org/10.1021/bc200090x |Bioconjugate Chem. 2011, 22, 1545–1558
Bioconjugate Chemistry
ARTICLE
Figure 8. PEGylation reaction for lysozyme and 5 kDa PEG at a pH of 4 (top left) and at a pH of 7 (top right), scFv, and 5 kDa PEG at pH 4 (bottom left) and 7 (bottom right). For lysozyme PEGylation, 21 °C, 20 h reaction time, 2-fold PEG excess, 1 g/L protein; and for scFv, 1 g/L protein, 21 °C, 20 h reaction time, 5-fold PEG excess. Experimental data shown as crosses, simulated data shown as straight lines. Green, unmodified protein; red, MonoPEG-protein; blue, Di-PEG-protein; yellow, Tri-PEG-protein.
The reaction rates r were thereby defined r 1 ¼ PEGreacting protein k1 r 2 ¼ PEGreacting Mono-PEG-protein k2 r 3 ¼ PEGreacting Di-PEG-protein k3 r 4 ¼ PEGreacting k4 and the differential equations were specified dxðPEGall Þ ¼ r 1 r 2 r 3 dxðproteinÞ ¼ r 1 dxðMono-PEG-proteinÞ ¼ r 1 r 2 dxðDi-PEG-proteinÞ ¼ r 2 r 3 dxðTri-PEG-proteinÞ ¼ r 3 dxðPEGreacting Þ ¼ r 1 r 2 r 3 r 4 dxðPEGinactive Þ ¼ r 4
The differential equations were solved as described above. The results of the kinetic simulation now fitted the experimental data very well. For lysozyme, an excellent match at all pH values could be achieved, and for scFv, the simulated data fitted the experimental data relatively good (Figure 8). All experimental data sets were simulated with the above demonstrated set of differential equations. The rate constants k1 and k2 were determined and are listed for different temperatures in Tables 3 and 4. The rate constant k1 was always higher than k2 in the case of lysozyme, indicating that the Mono-PEGylation was much faster than the addition of a second PEG. In contrast, the rate constant k2 exceeded k1 in an acidic pH and at very high protein concentrations in the case of the PEGylation of scFv with a 5 kDa PEG. However, using the 30 kDa PEG the rate constant k1 exceeded k2 in the case of both scFv and lysozyme (data not shown). The calculated rate constants k1 for the reaction of protein to MonoPEGylated protein were in the range 0.0020.016 L/mol min for lysozyme and 0.0090.3 L/mol min for scFv. The rate constants k2 generally were about half of the calculated values for k1. The rate constants should behave constant at different protein concentrations; this could be observed for lysozyme PEGylation with both PEG sizes. However, when using different protein 1551
dx.doi.org/10.1021/bc200090x |Bioconjugate Chem. 2011, 22, 1545–1558
Bioconjugate Chemistry
ARTICLE
Table 2. Relative Migration Length (Rf), Theoretical Sizes, and Calculated Sizesa for Analyzed Samples of PEGylated scFv theoretical apparent size of Rf [cm]
sample
theoretical size [kDa]
calculated size [kDa]
PEGylated protein in SDS-PAGE [kDa]
1.75
Di-PEG-30-scFv
88.5
156.82
128.5
2.1
PEG 30 kDa Dimer
60
116.01
100
2.6
Mono PEG-30-scFv
58.5
81.50
78.5
3.5
PEG 30 kDa
30
49.86
50
3.25
Tri-PEG-5-scFv
43.5
56.36
52.5
3.7
Di-PEG-5-scFv
38.5
45.48
44.5
4.25 5
PEG-5-scFv scFv
33.5 28.5
36.17 27.65
36.5 28.5
a
Column 4: calculated with the protein marker only. Column 5: calculated theoretical apparent size, taking into account the calculated size of PEG as it appeared in SDS-PAGE.
Table 3. PEGylation of Lysozyme at Different Temperatures and the Resulting Rate Constants k1, k2, and Yield, Conversion, and Selectivity after 20 h Reaction Time temp [°C ]
PEG [kDa]
k1 [L/mol min]
k2 [L/mol min]
yield
10
5
0.0037
0.0017
0.34
0.36
0.94
21
5
0.0125
0.0061
0.5
0.66
0.76
30 10
5 30
0.0165 0.0028
0.0089 0.0013
0.49 0.29
0.68 0.29
0.72 1.00
21
30
0.0075
0.003
0.43
0.45
0.96
30
30
0.0124
0.0022
0.51
0.59
0.86
conversion
selectivity
Table 4. PEGylation of scFv at Different Temperatures and the Resulting Rate Constants k1, k2, and Yield, Conversion, and Selectivity after 20 h Reaction Time temp [°C ]
PEG [kDa]
k1 [L/mol min]
k2 [L/mol min]
yield
conversion
selectivity
10 20
5 5
0.0199 0.0144
0.0101 0.0086
0.56 0.48
0.73 0.61
0.77 0.79
30
5
0.0727
0.0424
0.43
0.87
0.50
10
30
0.0153
0.0071
0.52
0.67
0.78
20
30
0.0115
0.0011
0.45
0.50
0.91
30
30
0.0657
0.0312
0.53
0.86
0.61
concentrations of scFv a wide difference in the rate constants was obtained (data not shown). An explanation could be that the evaluation of the chromatograms obtained at very high scFv concentrations was difficult due to a poor separation. In addition, it could be possible that the chosen reaction mechanism did not describe the reaction perfectly. Especially at high protein concentrations, the formation of protein aggregates could be a problem for peak evaluation and the reaction model. The influence of temperature on reaction velocities was very high. At 30 °C, the PEGylation reaction for lysozyme was fastest, and at a temperature of 10 °C, it was slowest. For scFv, this behavior changed; the reaction at 10 °C was faster than at 20 °C. In theory, a temperature rise should increase the reaction velocity.30 For lysozyme, with 5 and 30 kDa PEG the reaction velocity doubled with a temperature increase of 10 °C. For example, a reaction with 30 kDa PEG and lysozyme at 10 °C had a k1 of 0.0028 L/mol min; at 21 °C, k1 was 0.0063 L/mol min, and at 30 °C, 0.0124 L/mol min. For scFv, the 20 °C value was lower than the 10 °C value, whereas the 30 °C value was nearly four times the 10 °C value. Thus, there seems to be a problem in the mid-temperature range. The reaction did not follow the usual
standards in this temperature range (Tables 3 and 4). The modeling of the scFv PEGylation thereby showed two problems, one was the dependence of the reaction rate constant on the scFv concentration, which indicates that the reaction is influenced by a different mechanism and might not follow a reaction of first order as lysozyme did. The second problem was the temperature dependent behavior. Production Parameters. To decide which reaction conditions should be chosen for a scale-up, the conversion, yield, and selectivity of each reaction were monitored after 5, 10, and 20 h and plotted against different ambient conditions such as the pH ranging from 4 to 7, the temperature from 10 to 30 °C as well as the protein concentration from 1 to 10 g/L and the PEG-to-protein ratio from 1- to 5-fold excess. All factors were analyzed individually, starting conditions are listed in the PEGylation of scFv section. Reaction Time. For all reactions, the selectivity slightly decreased over the time, whereas the yield and conversion increased continuously (Figure 9). pH. Yield increased with increasing pH for the 30 kDa version of scFv PEGylation, whereas the pH did not influence the lysozyme PEGylation significantly (data not shown). Likewise, 1552
dx.doi.org/10.1021/bc200090x |Bioconjugate Chem. 2011, 22, 1545–1558
Bioconjugate Chemistry
ARTICLE
Figure 9. Conversion, yield, and selectivity over the reaction time. Shown exemplarily for lysozyme PEGylation with 5 kDa PEG at standard conditions. Gray bars, yield; black bars, conversion; white bars, selectivity.
Figure 10. Yield over pH. Shown exemplarily for scFv PEGylation with 5 kDa PEG (left) and 30 kDa (right) at 21 °C after 20 h. Gray bars: yield.
Figure 11. Conversion, yield, and selectivity over the total protein concentration for scFv and 5 kDa PEG (left), 30 kDa PEG (mid left), and lysozyme with 5 kDa PEG (mid right) and 30 kDa PEG (right). Gray bars, yield; black bars, conversion; white bars, selectivity. The reactions were done at 21 °C, for lysozyme at a pH of 6, and for scFv, at pH 4, both with 20 h reaction time.
the scFv PEGylation with 5 kDa PEG was not affected significantly by pH changes (Figure 10). The increase in yield with increasing pH was congruent with the observation made for mPEG-aldehyde, namely, that inactivating side reactions decreased the more neutral the pH was (see mPEG-Aldehyde Inactivation). Regarding selectivity, a pH dependent trend was not visible (data not shown). Protein Concentration. An increase in total lysozyme concentration with unchanged PEG-to-protein ratio led to increased conversion and yield, whereas the selectivity decreased. A plateau was reached between 5 and 10 g/L protein (Figure 11). The increase in conversion at a constant PEG-to-protein ration was
also visible for scFv PEGylation; however, a plateau was not reached when 30 kDa PEG was used. For scFv, the yield decreased with increasing protein concentration whereas the conversion rose slightly for 5 kDa PEG and heavily for 30 kDa PEG (Figure 11). Selectivity decreased rapidly with increasing protein concentration for all proteins and PEGs. An explanation for the decrease in yield with rising scFv concentration is the massive decrease in selectivity leading to a reduced yield despite the rising conversion. PEG-to-Protein Ratio. The PEG-to-protein ratio was studied only for scFv PEGylation. With increasing PEG excess, the yield increased as did conversion. This could be observed using 30 kDa 1553
dx.doi.org/10.1021/bc200090x |Bioconjugate Chem. 2011, 22, 1545–1558
Bioconjugate Chemistry
ARTICLE
Figure 12. Conversion yield and selectivity over the PEG excess at constant 1 g/L scFv. For 5 kDa PEG (left), 30 kDa PEG (right). Gray bars, yield; black bars, conversion; white bars, selectivity. The reactions were done at 21 °C at a pH of 4 and 20 h reaction time.
Figure 13. Conversion, yield, and selectivity over the temperature at constant 1 g/L protein. For scFv and 5 kDa PEG (left), 30 kDa PEG (mid left). For lysozyme and 5 kDa PEG (mid right) and 30 kDa PEG (right). Gray bars, yield; black bars, conversion; white bars, selectivity. The reaction pH was 4 for scFv and 6 for lysozyme. The reaction time was 20 h.
PEG, whereas the use of 5 kDa PEG resulted in a plateau in yield between 5- and 10-fold PEG excess. The selectivity decreased with increasing PEG-to-protein ratio except for PEG 30 kDa where no difference between 1- and 5-fold PEG excess could be found (Figure 12). Temperature. The temperature rise for lysozyme PEGylation led to increased yield and conversion; a plateau was reached for PEG 5 kDa between 21 and 30 °C; however, the selectivity decreased with increasing temperature. Selectivity also decreased when the scFv was used, though a trend was not that clearly visible for all factors. The yield decreased for a PEGylation of scFv with 5 kDa PEG; using 30 kDa PEG, a trend was not visible (Figure 13). An irregularity in the temperature range of 20 °C could also be shown for the reaction rate constants of the scFv PEGylation (Table 4). A different mechanism could be responsible for the unexpected behavior of the reaction rate and yield, for example, an interaction between PEG and scFv. The temperature had no effect on the scFv PEGylation with 30 kDa PEG, whereas for 5 kDa PEG, the rising temperature led to a decrease of yield and selectivity. For lysozyme, an increase in yield and a decrease in selectivity was visible regardless of the PEG size. Best Conditions. The best condition for a PEGylation reaction is the best possible combination of high yield and high selectivity. An increasing reaction time increased yield as well as conversion independent of the conditions used. Of note, the selectivity was slightly reduced between 10 and 20 h reaction
time. Consequently, the reaction should be stopped before reaching a plateau. Indeed, a reaction time around 15 h guaranteed high yield and selectivity at the tested starting conditions. Regarding the pH, big differences were not visible; however, it could be shown (see mPEG-Aldehyde Inactivation) that mPEG-aldehyde gets inactivated at low pH values. Thereby, the best pH should be chosen with respect to the solubility of the protein. For example, solubility of lysozyme, which has a pI of 11, was not critical within the pH range used in this study. In contrast, the solubility of scFv, which has a pI in the neutral range, was strongly dependent on the pH value. It could be shown that scFv was more soluble in more acidic surroundings, which are therefore recommended for its PEGylation. However, studies exist, which discovered a pH dependency of the selectivity for certain PEGylation sites, especially at the N-terminus,31,32 that was not further investigated here. A rising protein concentration led to an increase in conversion though to a decrease in selectivity. The ideal protein concentration thus depends on the protein used. In the case of lysozyme, a concentration of 5 g/L was found to be a good compromise, whereas the ideal protein concentration of scFv was in the lower range and dependent on the PEG size. Long PEG chains require a higher protein concentration than short PEG chains. For example, 2 g/L was used for scFv PEGylation with 30 kDa PEG and only 1 g/L for PEGylation with 5 kDa PEG, because in the case of scFv PEGylation, with 5 kDa PEG an increase in protein concentration led to a decrease in yield. This can be 1554
dx.doi.org/10.1021/bc200090x |Bioconjugate Chem. 2011, 22, 1545–1558
Bioconjugate Chemistry
ARTICLE
Figure 14. PEGylation reaction of lysozyme (left) and scFv (right), exemplarily shown with a 5 kDa PEG (right) and a 30 kDa PEG (left). PEG was preincubated in reaction buffer containing 20 mM NaCNBH3 at a pH of 4 in the absence of protein. Continuous line, 0 min reaction time; dashed line, 4 h reaction time. No reaction occurred regardless the protein or PEG size.
explained by a lower solubility of scFv and possibly the formation of aggregates. Of note, upon PEGylation with a 30 kDa PEG the decrease in yield was not apparent. It is thus possible that long chain PEG increases the solubility of scFv. In addition, the concentration-dependent reaction rates of scFv point toward difficulties with the PEGylation reaction mechanism. For lysozyme, the reaction rates were concentration independent, as they should be with a working model, but for the scFv PEGylation, this was not the case. Here, an additional factor seems to disturb the expected reaction mechanism. The formation of aggregates of the hydrophobic scFv might have to be taken into account in future research. In addition, the temperature dependence of scFv PEGylation also did not show the expected behavior. In the case of lysozyme, an increase of the reaction rate with increasing temperature was determined as predicted by Arrhenius.27 The yield and reaction rate of scFv PEGylation at 30 °C were highest; however, at 10 °C these parameters were still higher than those obtained at 20 °C, which does not fit with the theory of van’t Hoff and Arrhenius.27 A different reaction mechanism could be responsible here, for example, an interaction between PEG and protein or an aggregation phenomenon could explain this protein concentration and temperature-dependent behavior, respectively. In addition, a hydrophobic force-driven interaction could easily be temperature dependent, leading to a higher interaction at lower temperature and a better reactivity at 10 °C because of a higher probability of PEG and protein to interact. The interaction could reach an end point at 30 °C prevented by the Brownian motion. A normal reaction progression is thus reached again. Apparently, this interaction would also influence the dependence of the reaction rate constant on the protein concentration, which showed an unusual behavior as already discussed above. Our suggestion would thus be using a reaction temperature of 15 °C for scFv PEGylation and 20 °C for lysozyme PEGylation. In the case of scFv PEGylation, a lower PEG-to-protein ratio led to higher selectivity but lower yield and conversion. A compromise would be to use a relatively high PEG-to-protein ratio such as 5:1. The optimized conditions were tested for lysozyme at a pH of 6 and a temperature of 20 °C with a 2-fold PEG excess and a
lysozyme concentration of 5 g/L. These conditions were chosen for both PEG sizes; only the reaction time varied. For PEGylation with 5 kDa, 15 h reaction time was chosen, whereas the PEGylation with 30 kDa proceeded for 20 h. Yield, conversion, and selectivity were very good for both PEG sizes. Especially, the yield could be improved to 0.54 for 5 kDa and 0.59 for 30 kDa PEG. Both values were the best ever tested and accompanied by a conversion of 0.7 and 0.6, respectively. The combination of optimal values for all parameters was an important improvement, which could also be approved for scFv PEGylation. Here, the yield could be increased to 0.6 and 0.57 for 5 and 30 kDa PEG, respectively, both together with excellent conversion and selectivity. mPEG-Aldehyde Inactivation. The inactivation reaction for mPEG-AL seems to be essential for mathematical modeling (see Kinetic Studies and Mathematic Modeling) even though mPEG-AL should not be susceptible to hydrolysis.14 At a neutral pH, inactivation of PEG did not occur; however, at acidic pH values inactivation took place. To analyze the inactivation process in more detail, mPEG-AL was incubated overnight in different buffers without the presence of protein followed by a PEGylation reaction the next day. It could be shown that the reaction did not take place with PEG preincubated in a NaCNBH3-containing buffer of pH 4 using neither lysozyme nor scFv as PEGylation protein. The size of the PEG molecule also did not affect the inactivation reaction (Figure 14).The hydrolysis of the imine can be ruled out or is only a part of the inactivation, as no protein and thereby no amino group was present during the inactivation overnight. It was not further investigated to determine what driving force or which chemical compound of the reaction mix play a role in the inactivation reaction. In the following experiments, the inactivation reaction was observed, as it also took place in the PEGylation reactions without preincubation. All factors mentioned above were observed. Especially, temperature and pH had a significant influence. Increasing temperature led to increased inactivation reaction rates though the factor of temperature dependent changes also seemed to be dependent on the protein. In the case of lysozyme, a change was visible within the low temperature 1555
dx.doi.org/10.1021/bc200090x |Bioconjugate Chem. 2011, 22, 1545–1558
Bioconjugate Chemistry
ARTICLE
the PEG functionalization even when the same PEGylation chemistry was used.28 This could also account for the different PEG inactivation rates for differently functionalized PEGs. This clearly shows the higher stability of mPEG-AL in comparison with mPEG-succinimidyl propionate.
Figure 15. Inactivation reaction rate constant k4 (gray bars) over the temperature, shown exemplarily for 5 kDa PEG with lysozyme and a PEG concentration of 0.07 mol/L at a pH of 6 (left) and scFv with a PEG concentration of 0.18 mol/L at a pH of 4 (right).
Figure 16. Inactivation reaction rate constant k4 (gray bars) over the pH, shown for scFv and 30 kDa PEG at 21 °C at a constant PEG concentration of 0.18 mol/L.
ranges. The reaction rate remained constant between 21 and 30 °C. In the case of the scFv PEGylation, only small changes were visible within the low temperature regions; however, between 20 and 30 °C a considerable difference could be found (Figure 15). The temperature dependence was independent of the PEG size (data not shown). Nevertheless, it cannot be excluded that the pH is the factor influencing the absolute values of this experiment most, as the PEGylation of lysozyme was carried out at a pH of 6 whereas scFv-PEGylation took place at a pH of 4. In line with this, the pH dependence was clearly visible; the inactivation increased in acidic pH and decreased the more basic the environment got (Figure 16). This was visible for all proteins and all PEG sizes (Figure 16, shown exemplarily for scFv and 30 kDa PEG). The PEG concentration influenced the PEG inactivation slightly, increasing PEG concentration led to inactivation rate constants that are changed significantly only for the highest and lowest PEG-to-protein ratio. For example, the inactivation rate constant varied in the range 0.180.29 L/h between 1- and 10-fold excess of PEG in the case of scFv PEGylation with 5 kDa PEG. Buckley and co-workers calculated a rate constant of 1.31 L/h for mPEG-succinimidyl propionate inactivation.14 Of note, this inactivation rate is approximately 3-fold higher than the inactivation rate for mPEG-AL found at a pH of 3.2 in the present study. Hu et al. already found the PEGylation rate to be dependent on
’ CONCLUSIONS Lysozyme and a scFv were PEGylated successfully via aldehyde chemistry. The PEGylation reaction was monitored and a mathematical model was established. An additional inactivation reaction for mPEG-AL was included and the reaction could be simulated closely. In the case of lysozyme, an excellent working reaction model could be obtained, whereas in the case of scFv, the model did not work perfectly. By analyzing the reaction parameters such as reaction rates, conversion, yield, and selectivity, lysozyme PEGylation turned out to be a real model reaction. All factors behaved as expected; however, scFv PEGylation didn not display such an exemplary characteristic. Unfortunately, conditions used to obtain the best selectivity did not result in highest conversion. For example, the highest conversion was obtained at 30 °C; however, the selectivity was not good at such a high temperature. Furthermore, the reaction time appeared to be an important feature: The longer the reaction time, the better the yield and conversion, although selectivity decreased. Several studies describe that the pH is a crucial factor;6,8,13,15 however, we could only partly confirm this hypothesis. In our hands, PEGylation of lysozyme was pH-independent. However, PEGylation of scFv was improved at a neutral pH, although the solubility of the scFv is restricted to acidic pH values. Of note, the inactivation of mPEG-AL was found to be pH dependent. At acidic pH values, PEG inactivation is accelerated, and the pH should be chosen in the neutral or slightly acidic range, depending on the protein’s solubility behavior. Previous studies13,15 have described the isoform formation to be pH dependent; however, this will be investigated in future studies. The PEG-to-protein ratio was also examined. The lower it got, the more selective the reaction was. To obtain a compromise between selectivity and yield, the PEG excess should be approximately 2- to 5-fold. An ideal protein concentration of about 5 g/L for lysozyme was determined, whereas the scFv PEGylation requires a lower protein concentration dependent on the used PEG size. On the basis of the results obtained so far, optimal PEGylation conditions were obtained that resulted in a combination of high yield, conversion, and selectivity. The selected conditions used for lysozyme were 5 g/L protein concentration, 20 °C, 2-fold PEG excess, and a reaction time of 15 h for 5 kDa PEG and 20 h for 30 kDa PEG at a pH of 6. In the case of scFv PEGylation, the conditions used were a pH of 4, a reaction temperature of 15 °C, 15 h, and 5-fold PEG excess for both PEG sizes. Using 5 kDa PEG, a protein concentration of only 1 g/L was favorable, whereas in the case of 30 kDa PEG, 2 g/L was found to be the ideal concentration. Small mPEG-AL inactivation rates can be obtained by avoiding acidic pH values and high temperatures. The inactivation rate for m-PEG-AL could be calculated between 0.01 L/h and 0.5 L/h depending on environmental conditions. This is about a third of 1.31 L/h, the inactivation rate reported previously for mPEGsuccinimidyl propionate inactivation.14 MPEG-AL was found to be more stable than mPEG-succinimidyl propionate in the 1556
dx.doi.org/10.1021/bc200090x |Bioconjugate Chem. 2011, 22, 1545–1558
Bioconjugate Chemistry present study. The conducted experiments lead to the conclusion that the PEGylation reaction parameters need to be chosen and tested carefully for each protein and PEG size. Dependiong on the protein, the reaction mechanism seems to show slight differences. The presented model worked perfectly for the ideal lysozyme PEGylation, whereas the PEGylation of scFv was not simulated perfectly because of some unknown changes in reaction mechanism, which might be the formation of aggregates but were not further investigated here. Only minimal general predictions can be made, which are all defined by the mPEG-AL inactivation. Thus, high temperatures and acidic pH values should be avoided for aldehyde PEGylation.
’ AUTHOR INFORMATION Corresponding Author
*Anna Moosmann, Dipl. Biol. t.o., phone: +49 711 685 60262, e-mail:
[email protected]. Author Contributions #
Equal contributions.
’ ACKNOWLEDGMENT We thank the Bundesministerium fuer Bildung und Forschung (BMBF), Germany for supporting this work. ’ ABBREVIATIONS USED IEX, ion-exchange chromatography; mPEG-AL, methoxy-PEG-aldehyde; PEG, poly(ethylene glycol); Rf, relative migration length; scFv, single chain variable fragment; SEC, size exclusion chromatography ’ REFERENCES (1) Chowdhury, P. S., and Wu, H. (2005) Tailor-made antibody therapeutics. Methods 36, 11–24. (2) Chapman, A. P. (2002) PEGylated antibodies and antibody fragments for improved therapy: a review. Adv. Drug Delivery Rev. 54, 532–545. (3) Abuchowski, A., McCoy, J. R., Palczuk, N. C., van Es, T., and Davis, F. F. (1977) Effect of covalent attachment of polyethylene glycol on immunogenicity and circulating life of bovine liver catalase. J. Biol. Chem. 252, 3582–3586. (4) Veronese, F. M. (2001) Peptide and protein PEGylation: a review of problems and solutions. Biomaterials 22, 405–417. (5) Yang, K., Basu, A., Wang, M., Chintala, R., Hsieh, M., Liu, S., Hua, J., Zhang, Z., Zhou, J., Li, M., Phyu, H., Petti, G., Mendez, M., Janjua, H., Peng, P., Longley, C., Borowski, V., Mehlig, M., and Filpula, D. (2003) Tailoring structure-function and pharmacokinetic properties of single-chain Fv proteins by site-specific PEGylation. Protein Eng. 16, 761–770. (6) Zalipsky, S. (1995) Functionalized poly(ethylene glycol) for preparation of biologically relevant conjugates. Bioconjugate Chem. 6, 150–165. (7) Abuchowski, A., Van Es, T., Palczuk, N. C., and Davis, F. F. (1977) Alteration of immunological properties of bovine serum albumin by covalent attachment of polyethylene glycol. J. Biol. Chem. 252, 3578–3581. (8) Veronese, F. M. (2009) PEGylated protein drugs: Basic Science and Clinical Applications, Birkh€auser Verlag, Switzerland. (9) Fee, C. J., and Van Alstine, J. M. (2006) PEG-Proteins: Reaction engineering and separation issues. Chem. Eng. Sci. 61, 924–939. (10) Wang, Y., Youngster, S., Grace, M., Bausch, J., Bordens, R., and Wyss, D. F. (2002) Structural and biological characterization of pegylated recombinant interferon alpha-2b and its therapeutic implications. Adv. Drug Delivery Rev. 54, 547–570.
ARTICLE
(11) Monfardini, C., and Veronese, F. (1998) Stabilization of substances in circulation. Bioconjugate Chem. 9, 418–450. (12) Webster, R., Elliott, V., Park, B. K., Walker, D., Hankin, M., and Taupin, P. (2009) PEG and PEGconjugates toxicity. In PEGylated protein drugs: Basic Science and Clinical Applications. (Veronese, F. M., Ed.) pp 127147, Birkh€auser Verlag, Switzerland. (13) Wang J., Hu T., Liu, Y., Zhang, G., Ma, G., Su, and Z. (2010) Kinetic and stoichiometric analysis of the modification process for N-terminal PEGylation of staphylokinase. Anal. Biochem. doi:10.1016/ j.ab.2010.12.030 (14) Buckley, J. J., Finn, R. F., Mo, J., Bass, L. A., and Ho, S. V. (2008) PEGylation of biological macromolecules. In Process chemistry in the pharmaceutical industry, Vol. 2, Challenges in an ever changing climate (Gadamasetti, K., and Braish, T., Eds.) pp 383402, CRC Press, Boca Raton. (15) Nojima, Y., Iguchi, K., Suzuki, Y., and Sato, A. (2009) The pHdependent formation of PEGylated bovine lactoferrin by branched polyethylene glycol (PEG)-N-hydroxysuccinimide (NHS) active esters. Biol. Pharm. Bull. 32, 523–526. (16) Pabst, T. M., Buckley, J. J., Ramasubramanyan, N., and Hunter, A. K. (2007) Comparison of strong anion-exchangers for the purification of a PEGylated protein. J. Chromatogr., A 1147, 172–182. (17) Caserman, S., Kusterle, M., Kunstelj, M., Milunovic, T., Schiefermeier, M., Jevsevar, S., and Porekar, V. G. (2009) Correlations between in vitro potency of polyethylene glycolprotein conjugates and their chromatographic behavior. Anal. Biochem. 389, 27–31. (18) Moosmann, A., Christel, J., Boettinger, H., and Mueller, E. (2010) Analytical and preparative separation of PEGylated lysozyme for the characterization of chromatography media. J. Chromatogr., A 1217, 209–215. (19) M€uller, E., Josic, D., Schr€ oder, T., and Moosmann, A. (2010) Solubility and binding properties of PEGylated lysozyme derivatives with increasing molecular weight on hydrophobic-interaction chromatographic resins. J. Chromatogr., A 1217, 4696–4703. (20) Lee, H., and Park, T. G. (2003) A novel method for identifying PEGylation sites of protein using biotinylated PEG derivatives. J. Pharm. Sci. 92, 97–103. (21) Bornscheuer, U. T., Altenbuchner, J., and Meyer, H. H. (1999) Directed evolution of an esterase: screening of enzyme libraries based on pH-indicators and a growth assay. Bioorg. Med. Chem. 7, 2169–2173. (22) Khalameyzer, V., Fischer, I., Bornscheuer, U. T., and Altenbuchner, J. (1999) Screening, nucleotide sequence, and biochemical characterization of an esterase from Pseudomonas fluorescens with high activity towards lactones. Appl. Environ. Microbiol. 65, 477–482. (23) Rathore, A. S., Bilbrey, R. E., and Steinmeyer, D. E. (2003) Optimization of an osmotic shock procedure for isolation of a protein product expressed in E. coli. Biotechnol. Prog. 19, 1541–1546. (24) Laemmli, U. K. (1970) Cleavage of structural protein during assembly of the head bacteriophage T4. Nature 227, 680–685. (25) Bailon, P., Palleroni, A., Schaffer, C. A., Spence, C. L., Fung, W.-J., Porter, J. E., Ehrlich, G. K., Pan, W., Xu, Z.-X., Modi, M. W., Farid, A., and Berthold, W. (2001) Rational design of a potent, long-lasting form of interferon: A 40 kDa branched polyethylene glycol-conjugated interferon r-2a for the treatment of hepatitis C. Bioconjugate Chem. 12, 195–202. (26) Kurf€urst, M. M. (1992) Detection and molecular weight determination of polyethylene glycol-modified hirudin by staining after sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Anal. Biochem. 200, 244–248. (27) Hillebrand, U. (2009) St€ochiometrie: Eine Einf€uhrung in die € bungsaufgaben, Springer Verlag, Grundlagen mit Beispielen und U Berlin Heidelberg. (28) Hu, T., Prabhakaran, M., Acharya, S. A., and Manjula, B. N. (2005) Influence of the chemistry of conjugation of poly(ethylene glycol) to Hb on the oxygen-binding and solution properties of the PEG-Hb conjugate. Biochem. J. 392, 555–564. (29) Seely, J. E., and Richey, C. W. (2001) Use of ion-exchange chromatography and hydrophobic interaction chromatography in the preparation and recovery of polyethylene glycol-linked proteins. J. Chromatogr., A 908, 235–241. 1557
dx.doi.org/10.1021/bc200090x |Bioconjugate Chem. 2011, 22, 1545–1558
Bioconjugate Chemistry
ARTICLE
(30) Arnaut, L., Formosinho, S., and Burrows, H. (2007) Chemical kinetics: from molecular structure to chemical reactivity, Elsevier, Dordrecht. (31) Kinstler, O., Molineux, G., Treuheit, M., Ladd, D., and Gegg, C. (2002) Mono-N-terminal poly(ethylene glycol)protein conjugates. Adv. Drug Delivery Rev. 54, 477–485. (32) Lee, H., Jang, I. H., Ryu, S. H., and Park, T. G. (2003) N-terminal site-specific mono-PEGylation of epidermal growth factor. Pharm. Res. 5, 818–825.
’ NOTE ADDED AFTER ASAP PUBLICATION This paper was published on the Web on Aug 1, 2011, with errors in Table 2. The corrected version was reposted on Aug 17, 2011.
1558
dx.doi.org/10.1021/bc200090x |Bioconjugate Chem. 2011, 22, 1545–1558
