Article pubs.acs.org/ac
Detection of Histidine Oxidation in a Monoclonal Immunoglobulin Gamma (IgG) 1 Antibody Masato Amano,† Naoki Kobayashi,† Masayuki Yabuta,† Susumu Uchiyama,‡ and Kiichi Fukui*,‡ †
Biologics Technology Research Laboratories, Daiichi Sankyo Co, Ltd., 1-12-1, Shinomiya, Hiratsuka-shi, Kanagawa 254-0014, Japan Department of Biotechnology, Graduate School of Engineering, Osaka University, 2-1 Yamadaoka, Suita 565-0871, Osaka, Japan
‡
ABSTRACT: Although oxidation of methionine and tryptophan are known as popular chemical modifications that occur in monoclonal antibody (mAb) molecules, oxidation of other amino acids in mAbs has not been reported to date. In this study, oxidation of the histidine residue in a human immunoglobulin gamma (IgG) 1 molecule was discovered for the first time by mass spectrometry. The oxidation of a specific histidine located at the CH2 domain of IgG1 occurred under light stress, but it was not observed under heat stress. With the forced degradation study using several reactive oxygen species, the singlet oxygen was attributed to a reactive source of the histidine oxidation. The reaction mechanism of the histidine oxidation was proposed on the basis of the mass spectrometric analysis of IgG1 oxidized in deuterium oxide and hydrogen heavy oxide.
■
oxidation in IgG1 was finally confirmed by stress testing using the mAb treated with deuterium oxide (D2O) and hydrogen heavy oxide (H218O) and measuring the uptake of deuterium or heavy oxygen with mass spectrometry. Susceptibility of the histidine oxidation was also discussed from the relationship between the oxidation levels of the comprehensive histidine residues and the solvent accessibility surface areas (ASA) of their side chains. Half-lives and pKa values of the hydrogen and deuterium (H/D) at the C2position of the histidine were also investigated15 as another oxidation index. In addition, the reaction mechanism of methionine oxidation under heat stress was considered to be strictly different from that of light stress. As a source of ROS for the methionine oxidation, we focused on the generation of hydrogen peroxide from water catalyzed by tryptophan.16 The consensus tryptophan in the light chains of mAbs is known as a generator of hydrogen peroxide. 17 The oxidation mechanism of methionine under heat stress was also investigated using IgG1 and its mutant, of which consensus tryptophan was substituted for alanine.
INTRODUCTION Oxidation of methionine is a well-known chemical modification in monoclonal antibody (mAb) molecules.1−3 Oxidation of methionine affects not only the binding affinity of mAbs to antigens but also effector functions4,5 and serum half-life.6 For example, a decrease in the structural stability for the CH2 domain in an immunoglobulin G1 (IgG1) molecule has been reported as a consequence of methionine oxidation.7 Oxidation of tryptophan in mAbs has also been reported, and the tryptophan oxidations in the CDR regions might affect the binding activity to an antigen.8 Therefore, the oxidation levels of methionine or tryptophan are a major concern for the development of pharmaceutical mAbs. Although histidine is known as another oxidative species of the natural amino acids,9 oxidation of histidine in mAbs has not been reported yet. As summarized in the latest review about the characterization of mAbs,1 oxidation of methionine and tryptophan had been described only as chemical modifications that occur during purification, formulation, and storage processes. In this study, we surveyed the susceptibility of histidine residues to oxidation in a human IgG1 degraded under heat stress (40 °C/75% relative humidity (RH) for 3 months) and light stress (2000 l× of D65 illuminant/25 °C/65% RH for 25 days) with peptide mapping and mass spectrometry.10,11 Detection of the histidine oxidation in IgG1 has raised a question about the oxidation mechanism because it is quite important to control the oxidation levels of pharmaceutical mAbs. Protein oxidation can be induced by several reactive oxygen species (ROS), such as peroxide, singlet oxygen, hydroxyl radicals, and superoxide radical.12−14 To specify the ROS contributing to the histidine oxidation detected in the mAb, a forced degradation study of IgG1 with several ROS has been performed. The reaction mechanism of the histidine © 2014 American Chemical Society
■
EXPERIMENTAL SECTION Stress Testing. The monoclonal IgG1 antibody and its mutant were expressed in human embryonic kidney 293F cells, purified with a protein A column (GE Healthcare, Little Chalfont, UK) and ceramic hydroxyapatite column A (Bio-Rad Laboratories, Hercules, CA) and then formulated to PBS at a concentration of 20 mg/mL. The antibodies were stored at 40
Received: April 10, 2014 Accepted: June 18, 2014 Published: June 18, 2014 7536
dx.doi.org/10.1021/ac501300m | Anal. Chem. 2014, 86, 7536−7543
Analytical Chemistry
Article
Figure 1. HPLC−UV chromatograms of the digested peptide fragments for the (A) initial IgG1 and (B) IgG1 stressed under light exposure.
°C/75% RH for 3 months or 2000 l× (D65 illuminant)/25 °C for 25 days. For stress testing with stable isotope labeling,18 IgG1 was lyophilized and reconstituted with D2O and H218O prior to stress testing. Forced Oxidation. To 35 μL of carbonate bicarbonate buffer (pH 9.6), a 32 μL aliquot of 10 mM disodium molybdate(VI) dihydrate solution was added. Subsequently, an 8 μL aliquot of 0.2% H2O2 or 70% tert-butylhydroperoxide (tBHP) was added. After the addition of 5 μL of IgG1, the sample was incubated for 15 or 60 min for H2O2 oxidation or tBHP oxidation, respectively. Immediately after the incubation, the whole amount of the samples was applied to MicroSpin G25 columns (GE Healthcare), and the buffer was exchanged for 100 mM Tris−HCl buffer (pH 8.0) containing 0.02% polysorbate 80. Peptide Mapping. Into 100 μg of IgG1, denaturing buffer (8 M Gdn·HCl, 100 mM Tris−HCl, 0.02% polysorbate 80, pH 8.0) was added to make a volume of 86 μL. Reduction was carried out by adding 3 μL of 100 mM dithiothreitol and incubating it at 37 °C for 30 min. Alkylation was carried out with the addition of 7 μL of 100 mM iodoacetoamide and then incubation at room temperature for 15 min. Followed by the addition of 4 μL of dithiothreitol to quench the alkylation reaction, the buffer of the samples was exchanged with 100 mM Tris−HCl buffer (pH 8.0) containing 0.02% polysorbate 80 using MicroSpin G-25 columns. Digestion was carried out with the addition of 16 μL of 0.2 mg/mL modified trypsin and incubation at 37 °C for 30 min. The digestion reaction was stopped with the addition of 10 μL of 5% trifluoroacetic acid. The digested samples were separated by reversed-phase chromatography using an LC1200 system (Agilent Technologies, Palo Alto, CA) employing an AdvanceBio Peptide Map 2.1 × 150 mm, 2.7 μm column (Agilent Technologies). Mobile phase A contained a mixture of water and trifluoroacetic acid (1000:1), and mobile phase B contained a mixture of water, acetonitrile, and trifluoroacetic acid (400:3600:3). A linear gradient program of the mobile phase B from 0% to 45% was carried out over 120 min at a column temperature of 50 °C. The eluent was detected with a mass spectrometer, LTQ/XL Orbitrap (Thermo Fisher Scientific, Waltham, MA), equipped with an electrospray ion source in the positive ion mode for the m/z of 300−2000. For the MS/MS fragmentation analysis, the parent ions were fragmented with HCD at an isolation width of 6 Da and a collision energy of 35 V.
Deuterium Exchange of Histidine Side Chains. The buffer of IgG1 was exchanged to 150 mM NaCl using Vivaspin 15R (Sartorius AG, Göttingen, Germany). The sample was diluted up to 10-fold with buffers prepared from D2O as a diluent with different pHs (pH 4.0−9.5) and then incubated for 336 h at 25 °C. The buffers in D2O included 50 mM sodium acetate (pH 4.0 and 4.5), 50 mM MES (pH 5.0 to 7.0), and 50 mM HEPES (pH 7.5 to 9.5). The deuterium exchange ratios were evaluated by peptide mapping as described above with a minor modification of the linear gradient program of mobile phase B carried out from 0% to 52% over 30 min. Monitoring the changes in relative abundance between M and M + 1 isotopic peaks, pseudo-first-order rate constants of the deuterion exchange (kφ) were calculated as follows: kφ = ln⎡⎣1 + R336 − R 0⎤⎦/336
Here, R336 and R0 were intensity ratios of the M to M + 1 isotopic peaks after incubation in D2O for 336 h and that of the control sample, respectively. The kφ values were plotted against the pH, followed by the calculation of pKa and kφmax values as the inflection points and the upper asymptote of the 4parameter logistic fitting. The half-lives of the exchange reactions (t1/2) were calculated as follows:
t1/2 = ln 2/kφmax The full-length IgG1 model structure was generated using the crystal structure of human IgG against HIV-1 as a template (PDB ID: 1HZH) by Discovery Studio software 3.1 (Accelrys, San Diego, CA), followed by the calculation of solvent accessibility surface areas (ASA) of the side chains for methionine and histidine in the probe radius of 1.40.
■
RESULTS AND DISCUSSION Detection of Histidine Oxidation. As shown in Figure 1, novel peptide fragment peaks were observed at retention times of 74.8 and 75.6 min in the UV chromatogram when IgG1 was stressed under light exposure (2000 l×/25 °C for 25 days). The m/z’s of the novel peaks were 855.40 in doubly charged ions, which corresponded to +32 Da from the m/z of 839.40 detected as a doubly charged ion at the retention time of 72.7 min. Therefore, it was suspected that two oxygen atoms were incorporated into the native peptide fragment (m/z = 839.40) to produce the oxidized peptide fragment (m/z = 855.40).
7537
dx.doi.org/10.1021/ac501300m | Anal. Chem. 2014, 86, 7536−7543
Analytical Chemistry
Article
Figure 2. MS/MS spectra of (A) the native peptide fragment (m/z of 839.40) and (B) the oxidized peptide fragment (m/z of 855.40).
Table 1. Oxidation Levels of Methionine and Histidine under Heat and Light Stress m/za chain heavy chain
light chain
residue M34 M48 M81 H172 H208 H228 M256 H272 H289 H314 M362 M432e H433e H437e H439e M37 H193 H202
peptide fragments
oxidation (%)a oxidizedb
initialc
charge
native
heat
light
A24−K38 Q39−K63 A68−R87
2 2 2
927.929 62 1414.627 04 1074.030 46
935.927 08 1422.6245 1082.027 92
3.5 (1.8) 1.0(0.2) 1.4(0.4)
4.7 1.7 2.4
6.4 1.4 1.5
D152−K214
4
1547.76816d
1555.76562d
0.0 (0.0)
0.0
2.0
T227−K252 D253−R259 T260−K278 F279−K292 V306−K321 E360−K364
3 1 2 2 2 1
948.824 03 835.434 21 1070.017 35 839.404 63 904.506 89 637.286 15
959.487 31 851.429 12 1086.012 27 855.399 55 920.501 81 653.281 06
0.0 (0.0) 11.6 (5.4) 0.2 (0.4) 0.3 (0.4) 0.1 (0.5) 2.7(1.0)
0.0 45.4 0.3 0.2 0.0 14.4
0.2 41.0 3.9 10.6 2.4 15.5
W421−K443e
3
934.42722
Met: 939.75886; His: 945.09050
Met: 7.6 (0.7); His: 0.0 (0.0)
Met: 26.4; His: 0.0
Met: 26.1; His: 1.4
A25−R49 H193−K194 V195−K211
3
964.427 83
969.759 47
1.3
1.5
2
938.4671
954.462 02
1.3 (0.2) N/Af 0.0 (0.0)
0.0
0.0
Mass chromatograms were extracted for m/z ± 10 ppm to evaluate the oxidation ratio. b+15.99491 for methionine oxidation (account for O) and +31.98983 for histidine oxidation (account for O2). cMean results of the triplicate assay are shown. SD values are also shown in the parentheses. d The most abundant isotopic ion was used instead of the monoisotopic ion to improve the sensitivity. eOne residue of methionine and three residues of histidine were included in the pepdide fragment W421−K443. Each oxidation level of methionine and histidine was quantitated. fNot available because the peptide fragment was too short to retain the reversed-phase column. a
FNWYVDGVEVHNAK from the parent mass and the fragment ion series of b2, b3, y3, y4, and y5. The MS/MS spectra of the two oxidized peptide fragments were almost the
To reveal the structure of the peptide fragment peaks, MS/ MS spectra were obtained as shown in Figure 2. The native peptide fragment was annotated as the amino acid sequence of 7538
dx.doi.org/10.1021/ac501300m | Anal. Chem. 2014, 86, 7536−7543
Analytical Chemistry
Article
Table 2. Oxidation Levels of Methionine and Histidine with Forced Oxidation Reagents oxidation (%) chain
residue
heavy chain
M34 M48 M81 H172 H208 H228 M256 H272 H289 H314 M362 M432 H433 H437 H439 M37 H193 H202
light chain
control
H2O2
H2O2 + Mo
tBHP
Mo
2.7 1.0 1.9
3.1 1.1 2.2
2.1 1.5 2.4
3.6 1.3 2.1
75.5 71.2 58.8
0.0
0.0
0.0
0.0
0.0
0.0 11.2 0.2 0.4 0.0 2.0
0.0 28.5 0.2 0.4 0.0 11.7
1.8 99.9 10.7 28.8 0.2 100.0
0.0 12.0 0.2 0.4 0.0 2.5
0.0 89.2 0.3 0.3 0.0 37.9
Met: 7.7; His: 0.0
Met: 14.4; His: 0.0
Met: 41.0; His: 6.4
Met: 7.8; His: 0.0
Met: 90.2; His: 0.0
0.7
0.9
1.5
0.8
76.1
0.0
0.0
N/A 0.0
0.0
0.0
Figure 3. Mass spectra of oxidized His289-containing peptide stressed by light exposure in (A) water, (B) deuterium oxide, and (C) hydrogen heavy oxide. Mass spectra of the (D) native and (E) oxidized His289-containing peptide analyzed with the mobile phase using deuterium oxide.
was observed at m/z = 322.16 as well as the native peptide fragment. On the other hand, y41+ ions were detected differently
same, where the MS/MS spectrum of the front peak is shown in Figure 2B. For the oxidized peptide fragments, the y31+ ion 7539
dx.doi.org/10.1021/ac501300m | Anal. Chem. 2014, 86, 7536−7543
Analytical Chemistry
Article
Figure 4. Proposed pathways of histidine oxidation.
levels of the methionine were also evaluated, and an increase in the theoretical mass for the methionine oxidation was assumed to be approximately +16 Da. Oxidation of the histidine was observed specifically under light exposure, although the oxidation of methionine was observed under both heat and light stress. Therefore, it was suggested that the oxidation pathway of histidine is different from that of methionine and also that the oxidation of histidine takes place through a specific pathway of the light exposure. Reactive Oxidation Species for the Histidine Oxidation. Among several types of ROS, singlet oxygen is a candidate as a specific ROS generated by light exposure. To elucidate the involvement of ROS in histidine oxidation by light exposure, IgG1 was exposed to peroxide (H2O2), singlet oxygen (H2O2 with molybdate),20 or peroxyl and hydroxyl radicals (tBHP). As shown in Table 2, oxidations of the histidine were observed only when IgG1 was degraded with H2O2 with molybdate, indicating that the oxidated histidine was induced by singlet oxygen. Interestingly, oxidation levels of the methionine increased under degradation with H2O2 with molybdate, as compared with degradation with only H2O2. This result suggested that the methionine oxidation could be induced by peroxide and enhanced by singlet oxygen. Pathway of the Histidine Oxidation. To elucidate the oxygen source for the histidine oxidation, IgG1 was exposed to light under the coexistence of deuterium oxide (D2O) and hydrogen heavy oxide (H218O). Mass spectra of the peptide fragment containing oxidized His289 are shown in Figure 3. No mass shift of the monoisotopic ion was observed for the fragment of IgG1 stressed in D2O (panel B) compared with that in H2O (panel A), indicating no incorporation of the
between the native and oxidized peptide fragments. Instead of the y41+ ion at m/z = 469.25 for the native peptide fragment, y41+ ions at m/z = 483.23 were detected for the oxidized peptide fragments. Because m/z = 483.23 (+14 Da from the native histidine) corresponded to the dehydrated form (−18 Da) of the fragment ions containing doubly oxidized histidine (+32 Da), the oxidized peptide fragments were considered to contain oxidized histidine. The y51+ ions for the oxidized peptide fragments were also observed as a dehydrated form, indicating that the oxidized histidine was readily dehydrated in MS/MS analyses. It should be noted that, different from the elution profiles of peptide containing oxidized methionine and its nonoxidized counterpart, the peptide fragments containing oxidized histidine eluted later than the nonoxidized counterpart. The same elution profile for the histidine oxidation had been reported in a previous article. Angiotensin I with oxidized histidine was eluted later than the native Angiotensin I.19 Although doubly oxidized tryptophan has been reported for mAbs, the possibility of tryptophan oxidation was denied from the b31+ ions, which were observed at m/z = 448.20 both for the native and oxidized peptide fragments. Oxidation under Heat and Light Stress. Oxidation levels of the histidines in IgG1 were evaluated for both the heat- (40 °C/75% RH for 3 months) and light- (2000 l×/25 °C for 25 days stressed samples, as shown in Table 1. Oxidation levels were determined using the extracted ion chromatograms (XIC) of the native peptide fragments and the oxidized peptide fragments in which histidine residues are contained. The theoretical m/z of the peptide fragments containing oxidized histidine were calculated as approximately +32 Da from the native peptide fragments divided by the charges, assuming the incorporation of two oxygen atoms to histidine. Oxidation 7540
dx.doi.org/10.1021/ac501300m | Anal. Chem. 2014, 86, 7536−7543
Analytical Chemistry
Article
of the side chain for each histidine in IgG1 is shown in Table 3. Although there is an asymmetrical nature of IgG1 molecules in
hydrogen atoms from the water into the oxidized histidine. In contrast, an m/z shift of +1 was observed in the H218O (panel C) compared with that in H2O (panel A). Taking into account the doubly charged ion, this shift corresponded to a mass shift of +2 Da and the incorporation of only one heavy oxygen atom (18O) into the oxidized histidine because two oxygen atoms were involved in the oxidation of the histidine. These results suggested that one of the oxygen atoms originated from water but the other did not. Consequently, the histidine oxidation of IgG1 could be initiated with singlet oxygen that is generally generated from dissolved oxygen, followed by the additional oxidation mediated by the water molecule. Two pathways for the histidine oxidation have been proposed, as shown in Figure 4.21,22 Both schemes are initiated from singlet oxygen; however, hydration or adduct of hydration water are different. Although hydroxylated imidazolone could be produced in scheme A, water adducted imidazolone can be produced in scheme B. Water addition may occur at one of the three imidazolone carbons in scheme B. To clarify which scheme is the case for the histidine oxidation of IgG1, the mass spectrum of the peptide fragment containing His289 was measured employing D2O for the mobile phase of the analysis. Panel D in Figure 3 shows the mass spectrum of the native peptide fragment, and panel E shows that of the oxidized peptide fragment. The mass difference between the native and oxidized peptide was +33 Da. Taking into account the doubly charged ion, the m/z increase of +1 corresponded to the mass shift of +2 Da for the histidine oxidation analyzed in D2O compared with that in H2O. Thus, the incorporation of two dissociative protons into the imidazole group of the histidine by oxidation was indicated. For scheme A, one hydroxyl group and one secondary amine at the hydroxylated imidazolone could be suitable for the dissociative proton. In contrast, there were no dissociative protons in the wateradducted imidazolone of scheme B. Adducted water might have substituted D2O during the analysis; however, this possibility was denied because heavy oxygen remained during the analysis, as shown in panel C of Figure 3. As a result, we can conclude that scheme A is the histidine oxidation pathway of IgG1. We did not observe any intermediate product of the oxidation pathway in the light-exposed IgG1 by peptide mapping. Therefore, it was suggested that the intermediate products were processed promptly to produce hydroxylated imidazolone as the final product. Oxidation Site of Histidine. It has been suggested that the susceptibility of methionine in mAbs to oxidation depends to some extent on the level of the surface exposure of each methionine to solvent.3 As another example of the susceptibility to oxidation, it was reported that oxidative carbanion sites in an IgG1 were located on the surface of the protein and accessible to the solvent.24 As summarized in Table 1, significant levels of methionine oxidation were observed in our study at Met256 and Met432, which are located close to the CH2 and CH3 domain interface of the IgG1 heavy chain and are moderately exposed to the solvent as previously reported. The percent solvent accessibility surface area (ASA) of the side chains for Met256 and Met432 were 35% and 8%, respectively; however, ASA is not an accurate predictor of methionine oxidation, mainly because of the potential structural differences between the crystal structure and solution structure, which involves dynamics of proteins.7 The susceptibilities of histidine to oxidation were also investigated in terms of the surface exposure. The ASA percent
Table 3. Oxidation Levels of Methionine and Histidine with Forced Oxidation Reagents chain heavy
light chain a
residue
ASA (%)
pKa
ti/2 (day)
H172 H208 H228 H272 H289 H314 H433 H437 H439 H193 H202
18 0 92 40 93 35 0 100 50 44 0
8.2
49a
6.4 7.0 6.2
19 19 44
b
b
7.2
12a
N/Ac b
b
b
Average of the two or three histidine residues. Half lives cannot be evaluated because H/D exchange was not observed up to pH 9.5. cNot available.
crystal structures,23 the mean ASA percents of two heavy and light chains were calculated because the symmetrical dynamic structures of IgG1 are predicted in solution. Half-lives and pKa values of the hydrogen and deuterium (H/ D) at the C2-position of the histidine were also investigated as another index for the oxidation. The half-life and pKa values of each histidine were calculated from the titration plots of the first-order rate constants of the deuterium exchange (kφ) against the pH shown in Figure 5.
Figure 5. pH Dependence of pseudo-first-order rate constants for H/ D exchange at the C2-position of the histidine residues.
Oxidation of the histidine residues in IgG1 was observed predominantly at His289 in Table 1, as described above, and the ASA percent of the His289 side chain was 93%. However, significant oxidation could not be observed at His228 or His437, even though the ASA percents of these side chains are comparable to that of His289. These results indicate that, although solvent accessibilities could be one index of the histidine oxidation, other factors are probably related to the oxidation as well as methionine. His289 had a longer half-life of H/D exchange at the C2position than that of His228, even though the ASA percents of 7541
dx.doi.org/10.1021/ac501300m | Anal. Chem. 2014, 86, 7536−7543
Analytical Chemistry
Article
factors that induce oxidation, such as the interaction of amino acid residues around the histidine, according to the results of H/D exchange study at the C2-position of histidine. The oxidation of histidine in IgG1 is initiated by singlet oxygen, followed by hydration to produce hydroxylated imidazolone. Therefore, it is suggested that oxidation may be suppressed by a decrease in the dissolved oxygen concentration that is a source of singlet oxygen. Although methionine was oxidized under both heat and light stress, the reactive oxygen source of the heat stress was considered to be not singlet oxygen but peroxide, which is generated from dissolved oxygen catalyzed by consensus tryptophan in the light chain. It was suggested that the oxidation under heat stress may also be suppressed by a decrease in the dissolved oxygen concentration.
both residues are very close to each other, suggesting the interaction of His289 to the amino acid residues located around His289. Therefore, it was concluded that the particular interaction between His289 and the surrounding amino acids could accelerate the oxidation by chemically catalyzing the reaction or changing the orientation or the motility of the side chain. The dissociation status of the histidine side chain would not affect the susceptibility of the oxidation because we could not confirm any relationship between the oxidation levels and pKa. Oxidation Pathway of Methionine under Heat Stress. Although the oxidation of histidine could not be observed under heat stress, the oxidation of methionine was observed. The reaction mechanism of methionine oxidation under heat stress was considered to be significantly different from that under light stress; hence, it was not derived from singlet oxygen. For the mAb molecules, it has been shown that the consensus tryptophan in light chain framework 2 of variable regions (W35 in Kabat numbering, W39 for our IgG1) can catalyze the oxidation of water and generate hydrogen peroxide.17 Thus, peroxide might most likely be the reactive oxygen species instead of singlet oxygen under heat stress. To prove this, we carried out heat stress testing of W39A mutant IgG1 in which the tryptophan as a substrate generator for the oxidation was substituted by alanine. The oxidation levels of the methionine (Met256, Met362, and Met432) in the Fc domains of the W39A mutant were completely suppressed compared with that of the wild type, as shown in Table 4.
■
*Phone: +81-6-6879-7440. Fax: +81-6-6879-7441. E-mail:
[email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The authors highly acknowledge Jun Hasegawa, Chihiro Hirose, Toshi Kajiro, Makoto Ono, and Takefumi Kawabe for their kind support and helpful discussions.
■
Table 4. Oxidation Levels of Methionine for W39A-Mutated IgG1 under Heat Stress chain
residue
wild type, initial
wild type, heat
W39A, heat
M256 M362 M432
11.6 2.7 7.6
45.4 14.4 26.4
12.5 2.8 6.2
REFERENCES
(1) Beck, A.; Wagner-Rousset, E.; Ayoub, D.; Van, D. A.; SanglierCianférani, S. Anal. Chem. 2013, 85, 715−36. (2) Liu, H.; Gaza-Bulseco, G.; Zhou, L. J. Am. Soc. Mass. Spectrom. 2009, 20, 525−528. (3) Chumase, C.; Gaza-Bulseco, G.; Sun, J.; Liu, H. J. Chromatogr., B 2007, 850, 285−294. (4) Kuo, T. T.; Aveson, V. G. mAbs 2011, 3, 422−430. (5) Lam, X. M.; Yang, J. Y.; Cleland, J. L. J. Pharm. Sci. 1997, 86, 1250−1255. (6) Wang, W.; Vlasak, J.; Li, Y.; Pristatsky, P.; Fang, Y.; Pittman, T.; Roman, J.; Wang, Y.; Prueksaritanont, T.; Ionescu, R. Mol. Immunol. 2011, 48, 860−866. (7) Liu, D.; Ren, D.; Huang, H.; Dankberg, J.; Rosenfeld, R.; Li, L.; Brems, D. N.; Remmele, R. L., Jr. Biochemistry 2008, 47, 5088−5100. (8) Yang, J.; Wang, S.; Liu, J.; Raghani, A. J. Chromatogr., A 2007, 1156, 174−182. (9) Chang, S. H.; Teshima, G. M.; Milby, T.; Gillece-Castro, B.; Canova-Davis, E. Anal. Biochem. 1997, 244, 221−227. (10) Ren, D.; Pipes, G. D.; Liu, D.; Shih, L. Y.; Nichols, A. C.; Treuheit, M. J.; Brems, D. N.; Bondarenko, P. V. Anal. Biochem. 2009, 392, 12−21. (11) Dick, L. W., Jr.; Mahon, D.; Qiu, D.; Cheng, K. C. J. Chromatogr., B 2009, 877, 230−236. (12) Junyan, A. J.; Boyan, Z.; Wilson, C. Y.; John, W. J. Pharm. Sci. 2009, 98, 4485−4500. (13) Davies, M. J. Biochem. Biophys. Res. Commun. 2003, 305, 761− 770. (14) Barnes, S.; Shonsey, E. M.; Eliuk, S. M.; Stella, D.; Barrett, K.; Srivastava, O. P.; Kim, H.; Renfrow, M. B. Biochem. Soc. Trans. 2008, 36, 1037−1044. (15) Miyagi, M.; Nakazawa, T. Anal. Chem. 2008, 80, 6481−6487. (16) Sreedhara, A.; Lau, K.; Li, C.; Hosken, B.; Macchi, F.; Zhan, D.; Shen, A.; Steinmann, D.; Schöneich, C.; Lentz, Y. Mol. Pharmacol. 2013, 10, 278−288. (17) Wentworth, P., Jr.; Jones, L. H.; Wentworth, A. D.; Zhu, X.; Larsen, N. A.; Wilson, I. A.; Xu, X.; Goddard, W. A., 3rd; Janda, K. D.; Eschenmoser, A.; Lerner, R. A. Science 2001, 293, 1806−1811.
oxidation (%) heavy chain
AUTHOR INFORMATION
Corresponding Author
Oxidation of histidine was not detected for W39A under heat stress as well as in the wild type. Therefore, we concluded that the methionine oxidation under heat stress was induced by the hydrogen peroxide generated from dissolved oxygen catalyzed by the consensus tryptophan in the light chains.
■
CONCLUSION Oxidation of the histidine was detected for the first time in IgG1 when stressed by light exposure. Because the significantly oxidized histidine was located in the Fc region of the antibody, this modification could have occurred for all the monoclonal IgG1 antibodies. Revealing a novel potential oxidation site provided possibilities to analyze mAbs more strictly and easily in the future. For example, histidine oxidation can be evaluated more easily by the middle-down analysis with electron transfer dissociation by considering the histidine residue as a possible oxidation site.25 The knowledge about the histidine oxidation is also valuable for the development of biosimilar mAbs or Fcfusion proteins because the histidine oxidation could be a critical quality attribute that impacts efficacy or safety.26,27 The high solvent accessibility of the histidine might be a cause of the oxidation; however, it was difficult to predict the oxidation level accurately by the ASA calculated from the crystal ray structures, which did not reflect the dynamics of proteins in solution. Furthermore, there are other possible 7542
dx.doi.org/10.1021/ac501300m | Anal. Chem. 2014, 86, 7536−7543
Analytical Chemistry
Article
(18) Liu, H.; Ponniah, G.; Neill, A.; Patel, R.; Andrien, B. Anal. Chem. 2013, 85, 11705−11709. (19) Uchida, K.; Kawakishi, S. Arch. Biochem. Biophys. 1990, 283, 20− 6. (20) Boehme, K.; Brauer, H. D. Inorg. Chem. 1992, 31, 3468−3471. (21) Kim, J.; Rodriguez, M. E.; Guo, M.; Kenney, M. E.; Oleinick, N. L.; Anderson, V. E. Free Radic. Biol. Med. 2008, 44, 1700−1711. (22) Au, V.; Madison, S. A. Arch. Biochem. Biophys. 2000, 384, 133− 142. (23) Saphire, E. O.; Stanfield, R. L.; Crispin, M. D.; Morris, G.; Zwick, M. B.; Pantophlet, R. A.; Parren, P. W.; Rudd, P. M.; Dwek, R. A.; Burton, D. R.; Wilson, I. A. Adv. Exp. Med. Biol. 2003, 535, 55−66. (24) Yang, Y.; Stella, C.; Wang, W.; Schöneich, C.; Gennaro, L. Anal. Chem. 2014, 86, 4799−4806. (25) Fornelli, L.; Ayoub, D.; Aizikov, K.; Beck, A.; Tsybin, Y. O. Anal. Chem. 2014, 86, 3005−3012. (26) Beck, A.; Sanglier-Cianférani, S.; Van Dorsselaer, A. Anal. Chem. 2012, 84, 4637−4646. (27) Beck, A.; Reichert, J. M. mAbs 2011, 3, 415−416.
7543
dx.doi.org/10.1021/ac501300m | Anal. Chem. 2014, 86, 7536−7543
