Identification and Localization of Unpaired Cysteine Residues in

Jun 29, 2009 - The presence of free sulfhydryl groups in five recombinant monoclonal antibodies and their locations were further investigated in the c...
0 downloads 0 Views 3MB Size
Download PDF
Anal. Chem. 2009, 81, 6449–6457

Identification and Localization of Unpaired Cysteine Residues in Monoclonal Antibodies by Fluorescence Labeling and Mass Spectrometry Chris Chumsae, Georgeen Gaza-Bulseco, and Hongcheng Liu* Process Sciences Department, 100 Research Drive, Abbott Bioresearch Center, Worcester, Massachusetts 01605 Each human IgG1 antibody contains a total of thirty-two cysteine residues. In theory, all of them are involved in disulfide bonds, and no free sulfhydryl should be detected. However, literature has suggested that the presence of low levels of free sulfhydryl groups is likely a common feature of recombinant and wild type IgG1 antibodies. Currently, there is little information correlating the presence of free sulfhydryl to specific cysteine residues. The presence of free sulfhydryl groups in five recombinant monoclonal antibodies and their locations were further investigated in the current study. Free sulfhydryl groups were first modified using 5-idoacetamidofluorescein (5-IAF), which was followed by reduction of disulfide bonds and alkylation with iodoacetic acid (IAA). This procedure allowed differentiation of free cysteine residues from cysteine residues that were involved in disulfide bonding. In addition, it allowed a sensitive fluorescence detection of peptides with free sulfhydryl groups. The locations of the free sulfhydryl groups were determined using mass spectrometry after fraction collection. The results indicated that different antibodies had different levels of free sulfhydryl due to multiple unpaired cysteine residues commonly in the constant domains. Furthermore, free sulfhydryl due to unpaired cysteine residues in the variable domains varied for different antibodies. Interestingly, free sulfhydryl was rarely associated with cysteine residues that were involved in interchain disulfide bonds. Immunoglobulin G subclass 1 (IgG1) is the major class of immunoglobulin in human serum, and the most common choice for the development of recombinant monoclonal therapeutics. A homogeneous antibody is encoded by a single gene for the light chain and a single gene for the heavy chain, which is converted into a heterogeneous population soon after synthesis due to multiple enzymatic and nonenzymatic modifications.1,2 Several modifications are related to the integrity of the disulfide bond structure including incomplete disulfide bond formation, disulfide bond scrambling, and breakage of disulfide bonds. * To whom correspondence is addressed. Phone: (508)849-2591. Fax: (508)7934885. E-mail: [email protected]. (1) Awdeh, Z. L.; Williamson, A. R.; Askonas, B. A. Biochem. J. 1970, 116, 241–248. (2) Liu, H.; Gaza-Bulseco, G.; Faldu, D.; Chumsae, C.; Sun, J. J. Pharm. Sci. 2008, 97, 2426–2447. 10.1021/ac900815z CCC: $40.75  2009 American Chemical Society Published on Web 06/29/2009

IgG1 consists of a total of thirty-two cysteine residues, which form sixteen disulfide bonds. Twelve of them are intrachain disulfide bonds that connect the two layers of antiparallel β-sheets of each domain, and four of them are interchain disulfide bonds that connect light chain and heavy chains.3,4 A diagram of a typical IgG1 antibody with the locations of cysteine residues and disulfide bond linkage is shown in Figure 1. In theory, there should be no free sulfhydryl. However, evidence suggested that human IgG1 most likely has low levels of free sulfhydryl. Less than 0.2 free sulfhydryl per mol of IgG under native and denaturing conditions using 1% SDS and 5 M guanidine hydrochloride was found in human IgG.5 It was determined that the low levels of free sulfhydryl were most likely associated with human IgG2 but not IgG1.6,7 On the other hand, one accessible free sulfhydryl under native conditions and a total of four free sulfhydryl in the presence of 10 M urea and 4% SDS in human IgG1 were detected.8 In a different study, 2.6 mol free sulfhydryl per mole of IgG1 from human serum was detected after incubation with 5 M guanidine hydrochloride at 50 °C for 2 h.9 Free sulfhydryl was also detected in recombinant monoclonal antibodies. Approximately 0.03 and 0.08 mol free sulfhydryl per mole of recombinant IgG1 monoclonal antibodies were detected under native conditions and in the presence of 5 M guanidine hydrochloride respectively.10 Yet, in another study, approximately 0.67 mol free sulfhydryl per mol of a recombinant IgG1 was detected after incubation at 100 °C in the presence of 1% SDS for 5 min.11 Indirect evidence also suggested the presence of free sulfhydryl. For example, multiple lower molecular weight bands of recombinant monoclonal antibodies on nonreducing SDS-PAGE10-12 and nonreducing SDSCE12 were likely formed due to disulfide bond scrambling triggered by the presence of low levels of free sulfhydryl. This hypothesis was supported by the observation that cysteine as a (3) (4) (5) (6) (7)

(8) (9) (10) (11) (12)

Padlan, E. A. Mol. Immunol. 1994, 31, 169–217. Padlan, E. A. Adv. Protein Chem. 1996, 49, 57–133. Cecil, R.; Stevenson, G. T. Biochem. J. 1965, 97, 569–572. Schauenstein, E.; Dachs, F.; Reiter, M.; Gombotz, H.; List, W. Int. Arch. Allergy Appl. Immunol. 1986, 80, 174–179. Schauenstein, E.; Schauenstein, K.; Dachs, F.; Reiter, M.; Leitsberger, A.; Weblacher, M.; Maninger, K.; Horejsi, H.; Steinschifter, W.; Hirschmann, C.; Felsner, P. Biochem. Mol. Biol. Int. 1996, 40, 433–446. Gevondyan, N. M.; Volynskaia, A. M.; Gevondyan, V. S. Biochemistry (Moscow) 2006, 71, 279–284. Lacy, E. R.; Baker, M.; Brigham-Burke, M. Anal. Biochem. 2008, 382, 66–68. Zhang, W.; Czupryn, M. J. Biotechnol. Prog. 2002, 18, 509–513. Liu, H.; Gaza-Bulseco, G.; Chumsae, C.; Newby-Kew, A. Biotechnol. Lett. 2007, 29, 1611–1622, Epub 2007 Jul 1614. Hunt, G.; Nashabeh, W. Anal. Chem. 1999, 71, 2390–2397.

Analytical Chemistry, Vol. 81, No. 15, August 1, 2009

6449

Figure 1. A diagram of a typical IgG1 antibody. Only half of the molecule is shown. The domains are labeled as VL for the light chain variable domain, CL for the light chain constant domain, VH for the heavy chain variable domain, CH1 for the heavy chain constant domain 1, CH2 for the heavy chain constant domain 2, CH3 for the constant domain 3, and hinge for the hinge region. Each domain has two cysteine residues, which are labeled as 1 and 2 from the N-termini of the light chain or the heavy chain. Hinge region has one cysteine residue from the light chain and three from the heavy chain, which are labeled as 1, 2, and 3.

mild reducing reagent promoted the formation of the lower molecular weight bands, while N-ethylmaleimide, which reacts with free sulfhydryl, decreased the levels of the lower molecular weight bands.11 Unpaired cysteine residues and breakage of disulfide bonds are two major sources of free sulfhydryl groups. First, disulfide bonds may not be completely formed during antibody synthesis and assembly. Distinct peaks were observed when a recombinant monoclonal antibody was analyzed by hydrophobic interaction chromatography, which was determined to contain antibody with unpaired Cys 22 and Cys 96 of the heavy chain variable domain.13 Formation of the disulfide bond between these two cysteine residues was improved by adding copper sulfide to the cell culture medium.14 Second, free sulfhydryl can be produced by breakage of disulfide bonds through β-elimination, which occurs under basic conditions.15-19 Presumably, free sulfhydryl due to incomplete disulfide bond can occur in cysteine residues that are located in any domain. On the other hand, β-elimination most likely results in free sulfhydryl that are involved in interchain disulfide bonds because only these disulfide bonds are exposed and accessible to solvents.4,20-22 Disulfide bond formation is critical for protein folding, assembly, structure, stability, and function. The recombinant monoclonal antibody with unpaired cysteine residues in the heavy chain variable domain showed only about 40% potency compared to antibody with the respective intact disulfide bonds.13 Higher levels of free sulfhydryl resulted in decreased thermal stability in multiple antibodies.9 Thus, it is important to determine the sites of free (13) Harris, R. J. Dev. Biol. (Basel, Switz.) 2005, 122, 117–127. (14) Chaderjian, W. B.; Chin, E. T.; Harris, R. J.; Etcheverry, T. M. Biotechnol. Prog. 2005, 21, 550–553. (15) Nashef, A. S.; Osuga, D. T.; Lee, H. S.; Ahmed, A. I.; Whitaker, J. R.; Feeney, R. E. J. Agric. Food Chem. 1977, 25, 245–251. (16) Florence, T. M. Biochem. J. 1980, 189, 507–520. (17) Galande, A. K.; Trent, J. O.; Spatola, A. F. Biopolymers 2003, 71, 534–551. (18) Tous, G. I.; Wei, Z.; Feng, J.; Bilbulian, S.; Bowen, S.; Smith, J.; Strouse, R.; McGeehan, P.; Casas-Finet, J.; Schenerman, M. A. Anal. Chem. 2005, 77, 2675–2682. (19) Cohen, S. L.; Price, C.; Vlasak, J. J. Am. Chem. Soc. 2007, 129, 6976–6977, Epub 2007 May 6915. (20) Amzel, L. M.; Poljak, R. J. Annu. Rev. Biochem. 1979, 48, 961–997. (21) Virella, G.; Parkhouse, R. M. Immunochemistry 1973, 10, 213–217. (22) Sears, D. W.; Mohrer, J.; Beychok, S. Biochemistry 1977, 16, 2031–2035. (23) Petrotchenko, E. V.; Pasek, D.; Elms, P.; Dokholyan, N. V.; Meissner, G.; Borchers, C. H. Anal. Chem. 2006, 78, 7959–7966.

6450

Analytical Chemistry, Vol. 81, No. 15, August 1, 2009

sulfhydryl in recombinant monoclonal antibodies, which can aid in the determination of the origin of free sulfhydryl. In the current study, the presence of free sulfhydryl groups and their locations in five recombinant monoclonal antibodies were investigated. Free sulfhydryl groups were labeled with 5-IAF, while disulfide bonded cysteine residues were alkylated using iodoacetic acid (IAA) after reduction. Modification by 5-IAF resulted in a significant molecular weight increase and thus enabled a simple detection of the free sulfhydryl groups by measuring antibody molecular weights. In addition, modification by 5-IAF allowed the detection of tryptic peptides that contained free sulfhydryl groups using HPLC with fluorescence detection, which facilitated fraction collection. The advantage of the overall strategy was demonstrated by determination of the distribution of free sulfhydryl groups in five recombinant monoclonal antibodies. MATERIALS AND METHODS Materials. Five recombinant monoclonal IgG1 antibodies (referred to as Mab-A, Mab-B, Mab-C, Mab-D, and Mab-E) were used in the current study. 5-Iodoacetamidofluorescein (5-IAF) was purchased from Pierce (Rockford, IL). Iodoacetic acid (IAA), dithiothreitol (DTT), and R-cyano-4-hydroxycinnamic acid were purchased from Sigma (St. Louis, MO). Acetonitrile, trifluoacetic acid (TFA), and guanidine hydrochloride were purchased from J. T. Baker (Phillipsburg, NJ). Formic acid was purchased from EMD (Gibbstown, NJ). N-Octylglucoside was purchased from Roche (Indianapolis, IN). PNGaseF was purchased from Prozyme (San Leandro, CA). Trypsin was purchased from Worthington (Lakewood, NJ). PNGaseF Digestion. One mg of each antibody at a concentration of 5 mg/mL in 200 µL of 100 mM Tris, pH 7.0 with 1% (w/v) N-octylglucoside was deglycosylated using 5 µL of PNGaseF and incubation at 37 °C for 18 h. N-Octylglucoside as a nonionic detergent was included in the sample preparation to facilitate PNGaseF digestion. Alkylation with 5-IAF. Optimal amounts of 5-IAF and guanidine hydrochloride were tested using Mab-A. Mab-A was diluted to a final concentration of 1 mg/mL in 100 mM Tris, pH 7, containing either 0, 2, 4, or 6 M guanidine hydrochloride. 5-IAF was added to aliquots of each sample to obtain final 5-IAF:antibody ratios of either 5:1, 10:1, or 50:1. The samples were incubated at room temperature for 2 h. Unreactive 5-IAF was removed from

Figure 2. Mass spectra of Mab-A modified in the presence of 0 M (A), 2 M (B), 4 M (C), or 6 M (D) guanidine hydrochloride using a 5-IAF to Mab-A ratio of 10:1.

the samples by buffer exchange to 100 mM Tris, pH 7.0 using NAP-10 columns by following the manufacture’s procedure (GE Healthcare, Piscataway, NJ). Levels of specific and nonspecific modifications were determined by measuring the intact molecular weights by LC-MS. The condition that resulted in the highest level of specific modification and lowest level of nonspecific modifications was considered optimal and used for further experiments. Antibodies A-D with and without PNGaseF digestion were modified by following the optimized procedure, which, as discussed later, used a 10:1 5-IAF:antibody molar ratio, 4 M guanidine hydrochloride in 100 mM Tris, pH 7.0, antibody concentration of 1 mg/mL, and incubation at room temperature for 2 h. LC-MS Determination of Antibody Intact Molecular Weights. An Agilent HPLC and a Q-TOF 6510 mass spectrometer (Agilent, Santa Clara, CA) were used to determine the molecular weights of the 5-IAF modified antibodies. Approximately 2 µg of each sample was loaded on to a protein microtrap (Michrom

Bioresources, Auburn, CA) at 95% mobile phase A (0.02% TFA and 0.08% FA in Milli-Q water) and 5% mobile phase B (0.02% TFA and 0.08% FA in acetonitrile). After 5 min, mobile phase B was increased to 95% within 1 min, maintained at 95% mobile phase B for 4 min, and then decreased to 5% mobile phase B in 1 min. The protein microtrap was equilibrated using 5% mobile phase B for 4 min before the next injection. The flow-rate was set at 50 µL/minute with the column oven set at 60 °C. The mass spectrometer was set at an m/z range of 600-3200, source temperature of 350 °C, and Vcap voltage of 4750 V. Trypsin Digestion. Antibodies modified by 5-IAF were denatured with 6 M guanidine hydrochloride in 100 mM Tris, pH 8.0 at room temperature for 15 min. The samples were then reduced with 10 mM DTT at 37 °C for 30 min and followed by alkylation with 25 mM IAA prepared in 1 M Tris, pH 8.0 at 37 °C for 30 min. The reduced and alkylated antibodies were then buffer exchanged to 10 mM Tris, pH 8.0. Trypsin was added to each Analytical Chemistry, Vol. 81, No. 15, August 1, 2009

6451

Figure 3. Mass spectra of Mab-A that was modified using a 5-IAF to antibody ratio of 5:1 (A), 10:1 (B), or 50:1 (C) in the presence of 4 M guanidine hydrochloride.

sample with a final ratio of a 1:20 trypsin:antibody ratio. Digestion was performed at 37 °C for 4 h. The peptides were analyzed by reversed-phase (RP) HPLC with a fluorescence detector. Peptide Mapping with Fluorescence Detection. A Shimadzu HPLC with a fluorescence detector and a Vydac C18 column (4.6 × 250 mm) were used to separate tryptic peptides and detect peptides that were modified with 5-IAF. Mobile phase A was 0.1% TFA in Milli-Q water. Mobile phase B was 0.1% TFA in acetonitrile. Approximately 100 µg of peptides from each antibody was loaded at 95% mobile phase A and 5% mobile phase B. After 5 min, mobile phase B was increased to 35% in 130 min. The column was washed with 95% mobile phase B for 5 min and then equilibrated using 5% mobile phase B for 10 min. Elution of 5-IAF modified peptides was monitored with an excitation wavelength of 490 nm and an emission wavelength of 520 nm. Fluorescence peaks were collected for further analyses. Matrix-Assisted-Laser-Desorption Ionization Time-ofFlight Mass Spectrometry (MALDI-TOF). Fluorescent peaks were analyzed using an AXIMA-CFR MALDI-TOF mass spectrometer (Shimadzu). Fractions were spotted onto a MALDI plate. Fractions containing low levels of peptides were concentrated using a speed-vacuum, C18-ziptips (Millipore, Billerica, MA), or both. After the samples were dried, 1 µL matrix was added to the spots on the MALDI plate. The matrix was prepared by dissolving 10 mg of R-cyano-4-hydroxycinnamic acid in 50% acetonitrile in Milli-Q water with 0.05% TFA. The laser power and scan ranges were optimized for each peptide. 6452

Analytical Chemistry, Vol. 81, No. 15, August 1, 2009

RESULTS AND DISCUSSION Principle of the Method. Specific modification of free sulfhydryl groups of cysteine residues with fluorescent dyes and subsequent analysis of fractions by mass spectrometry has been employed in identification of cysteine residues in proteins and peptides.23,24 The challenge in the current study was to differentiate low level of free sulfhydryl in antibodies that have multiple disulfide bonds. Antibodies were first modified using 5-IAF. Modification of one free sulfhydryl group by 5-IAF increased the antibody molecular weight by 387.4 D, which allowed the detection of free sulfhydryl groups by direct measurement of the molecular weights. In addition, modification by 5-IAF added a fluorescence tag to free sulfhydryl groups, which was used to differentiate from cysteine residues that were involved in disulfide bonds. Efficient removal of 5-IAF was critical for an accurate detection of free sulfhydryl groups. 5-IAF was first removed by buffer exchange using NAP-10 columns, which should remove the majority of 5-IAF. As will be discussed later, a 10:1 5-IAF:antibody ratio was used for the experiments, which required only 67.6 µM 5-IAF. Residual 5-IAF was expected to be negligible and should be quenched by an excess amount of DTT used for reduction of the disulfide bonds. Following the removal of 5-IAF, the antibodies were reduced and alkylated with iodoacetic acid for trypsin digestion. Trypsin digestion of a typical recombinant monoclonal IgG1 antibody commonly resulted in approximately sixty peptides (24) Chen, S. H.; Hsu, J. L.; Lin, F. S. Anal. Chem. 2008, 80, 5251–5259, Epub 2008 May 5231. (25) Boja, E. S.; Fales, H. M. Anal. Chem. 2001, 73, 3576–3582.

Figure 4. Mass spectra of antibodies A-E after modification with 5-IAF. Left panels represent native antibodies A-E. Right panels represent antibodies A-E after removal of oligosaccharides.

with a minimum of three amino acid residues that are likely retained by C18 column and detected by UV at 214 nm. However, only peptides that contained 5-IAF modified free sulfhydryl groups were detected by fluorescence detection, which facilitated fraction collection and further analysis by MALDI-TOF mass spectrometry. It thus enabled the comparison and assignment of the major sites of free sulfhydryl. Optimization of the Amount of Guanidine Hydrochloride and 5-IAF. The amount of guanidine hydrochloride and 5-IAF were optimized using Mab-A. Multiple peaks were observed in Mab-A due to the presence of various N-linked oligosaccharide structures and the presence or absence of a C-terminal lysine (Lys). The major glycoforms of recombinant monoclonal antibodies are core-fucosylated biantennary complex structures with either zero (G0), one (G1), or two (G2) terminal galactose residues. The heavy chain of the antibodies can exist with or without C-terminal Lys, which results in antibodies with either zero (Lys 0), one (Lys 1), or two (Lys 2) C-terminal Lys. Identities of the major peaks with various glycoforms and different numbers of C-terminal Lys are labeled in Figure 2A. Modification of one free sulfhydryl results in an average molecular weight increase of 387.4 Da. Reaction of 5-IAF with free sulfhydryl as a result of unpaired cysteine residues should result in molecular weight increments of 774.8 Da. The lack of peaks with a molecular weight increase of approximately 774.8 Da suggested that Mab-A was not modified in the absence of guanidine hydrochloride (Figure

2A) or in the presence of 2 M guanidine hydrochloride (Figure 2B). Peaks with molecular weights of approximately 148866 Da were observed when the antibody was modified in the presence of 4 M (Figure 2C) and 6 M (Figure 2D) guanidine hydrochloride. This molecular weight was approximately 783 Da higher than the main peak with the molecular weight of 148083 Da, which corresponded to Mab-A that was modified at two sites. The peaks with molecular weights of approximately 149028 Da were 162 Da higher than the peaks with two 5-IAF modifications, which corresponded to Mab-A also with two 5-IAF modifications with G1 instead of G0. No significant difference was observed in the presence of either 4 or 6 M guanidine hydrochloride, which suggested 4 M guanidine hydrochloride can efficiently expose all free sulfhydryl groups. The ratio of 5-IAF to Mab-A was also optimized in the presence of 4 M guanidine hydrochloride. Similar levels of Mab-A with two 5-IAF modifications were observed at molar ratios of 5:1 and 10:1 (Figure 3A,B). A more complicated mass spectrum was observed when Mab-A was modified using a 50:1 5-IAF:Mab-A ratio (Figure 3C). Peaks with the molecular weights of 148470 and 148632 Da corresponded to Mab-A (Lys 0 with either Gal 0 or Gal 1) with one 5-IAF, which suggested that using a 50:1 5-IAF:Mab-A ratio with an excess of the reagent resulted in nonspecific modifications with the Nterminal primary amine or the side chains of amino acids such Analytical Chemistry, Vol. 81, No. 15, August 1, 2009

6453

Figure 5. Chromatograms of tryptic peptides of antibodies A-E with fluorescence detection. Peaks a-f are common for antibodies A-E, and their identities are shown in Table 1. Peaks 1-5 vary for antibodies A-E, and their identities are shown in Table 2.

as lysine, tyrosine, and histidine as has been reported for iodoacetamide.25,26 High sequence similarities of different antibodies are expected to result in similar folding, structure, and stability. Therefore, the optimized condition of a 10:1 5-IAF:antibody ratio in the presence of 4 M guanidine hydrochloride was used for modification of all five antibodies. Detection of Free Sulfhydryl by LC-MS. The presence of free sulfhydryl was determined by measuring the molecular weights after 5-IAF treatment. Antibodies A-E with and without oligosaccharides were modified using the optimized conditions. Similar to the previous observation, multiple peaks were observed in antibodies A-E due to the presence of various N-linked oligosaccharide structures or a different number of C-terminal Lys or both (Figure 4 native, A-E). Multiple peaks in the spectra of antibodies A-E after removal of N-linked oligosaccharides were due to the presence of C-terminal Lys (Figure 4, deglycosylated, A-E). In addition, peaks with a molecular weight increase of approximately 776 Da, which corresponded to 5-IAF modification at two sites, were observed in the mass spectra of antibodies A-E before and after deglycosylation. The relative peak intensities of the antibodies with 5-IAF modification compared to the antibodies without 5-IAF modification were slightly different for different antibodies, which indicated different levels of free sulfhydryl among the different antibodies. Only peaks with two sites of 5-IAF modifications were observed suggesting that each modified molecule had two major unpaired cysteine residues or a single unformed disulfide bond. However, it was unclear whether the unpaired cysteine residues were at the same or different locations. It was also unclear whether antibodies A-E shared the same unpaired cysteine residues or difference in locations. Peptide mapping with fluorescence detec(26) Yang, Z.; Attygalle, A. B. J Mass Spectrom 2007, 42, 233–243. (27) Gadgil, H. S.; Bondarenko, P. V.; Pipes, G. D.; Dillon, T. M.; Banks, D.; Abel, J.; Kleemann, G. R.; Treuheit, M. J. Anal. Biochem. 2006, 355, 165– 174, Epub 2006 Jun 2015.

6454

Analytical Chemistry, Vol. 81, No. 15, August 1, 2009

Table 1. Identities of Peaks a-f molecular weights (MH+) peaks

observed

calculated

amino acid sequences

a b c d e f

637.0 1652.1 1491.5 2468.8 3130.8 2128.3

637.2 1651.8 1491.7 2469.1 3131.3 2128.0

CK STSGGTAALGCLVK NQVSLTCLVK TPEVTCVVVDVSHEDPEVK WQQGNVFSCSVMHEALHNHYTQK SGTASVVCLLNNFYPR

Table 2. Identities of Peaks 1-5 molecular weights (MH+) peaks

observed

calculated

amino acid sequences

1 2 3 4 5

855.2 1097.4 1079.2 2205.8 1497.4

855.3 1097.4 1079.5 2206.0 1497.7

MSCK VTMTCR VTITCR VYACEVTHQGLSSPVTK LSCAASGFNIK

tion was thus carried out to determine the distribution of free sulfhydryl groups. Peptide Mapping with Fluorescence Detection. Tryptic peptides of antibodies A-E were analyzed by RP-HPLC with fluorescence detection. Multiple peaks were observed in the fluorescence chromatograms of antibodies A-E (Figure 5), which suggested that the presence of free sulfhydryl groups was associated with multiple cysteine residues. In agreement with the previous observation by analyzing the intact molecular weights, it was clear from inspection of the fluorescence chromatograms that different antibodies had slightly different levels of free sulfhydryl. Peaks with similar retention times (peaks a-f) were observed in different antibodies suggesting free sulfhydryl groups in the same locations in different antibodies. In addition, peaks of different retention times (peaks 1-5) indicated differences among antibodies either in the

Figure 6. MALDI-TOF mass spectra of peaks a-f from Figure 5. Isotopic peaks are shown as insets. The data is summarized in Table 1.

locations of the free sulfhydryl or in amino acid sequences containing the same cysteine residue. Identification of the Sites of Free Sulfhydryl. The major peaks from peptide mapping were collected and analyzed by MALDI-TOF mass spectrometry to determine the locations of the free sulfhydryl. Identification was based on matching the observed peptide molecular weights with the calculated molecular weights of cysteine containing tryptic peptides from known amino acid sequences of the antibodies and the molecular weight increase due to the 5-IAF modification. Modification of a peptide with one free sulfhydryl groups resulted in a monoisotopic molecular weight increase of 387.1 Da. Peaks were classified into two groups. Peaks a-f were observed in all five antibodies, while peaks 1-5 were observed in a subset of antibodies. Peaks with the same retention times shared the same identities as analyzed by mass spectrometry. Representative mass spectra of peaks a-f from antibody-E are shown in Figure 6. The observed and theoretical molecular weights and the amino acid sequences of the 5-IAF modified peptides are summarized in Table 1. Representative mass spectra of peaks 1-5 are shown

in Figure 7. Identities of the peaks are summarized in Table 2. Peak 1 with a molecular weight of 855.2 Da was detected in both antibodies D and E, which corresponded to the common sequence of MSCK that was modified by 5-IAF. Peak 2 with a molecular weight of 1097.39 Da was only detected in antibody-E, which corresponded to the unique amino acid sequence of VTMTCR. Peak 3 was detected in antibodies A and C, which corresponded to the common amino acid sequence of VTITCR. Peak 4 was detected in antibodies A, B, C, and E, which corresponded to the amino acid sequence of VYACEVTHQGLSSPVTK. This amino acid sequence was also present in antibody-D; however, only a small peak was detected by fluorescence detection. Peak 5 was only detected in antibody-C, which corresponded to the unique amino acid sequence of LSCAASGFNIK. The peak that eluted immediately before peak 3 in antibody C had a molecular weight of 1664.7, which corresponded to the amino acid sequence of AEDTAVYYCSR. Locations of the Free Sulfhydryl Groups in the Antibodies. The locations of the free sulfhydryl groups in the domain structures of the recombinant monoclonal antibodies are sumAnalytical Chemistry, Vol. 81, No. 15, August 1, 2009

6455

Figure 7. MALDI-TOF mass spectra of peaks 1-5 from Figure 5. Isotopic peaks are shown as insets. The data is summarized in Table 2.

Table 3. Locations of Free Sulfhydryl in Antibodies A-Ea antibodies cysteine positions

Mab-1

Mab-2

Mab-3

Mab-4

Mab-5

VH-1 VH-2 CH1-1 CH1-2 Hinge-1 Hinge-2 Hinge-3 CH2-1 CH2-2 CH3-1 CH3-2 VL-1 VL-2 CL-1 CL-2 LC-hinge

ND ND +(peak b) ND ND ND ND +(peak d) +(peak a) +(peak c) +(peak e) +(peak 3) ND +(peak f) +(peak 4) ND

ND ND +(peak b) ND ND ND ND +(peak d) +(peak a) +(peak c) +(peak e) ND ND +(peak f) Low ND

+ (peak 5) + +(peak b) ND ND ND ND +(peak d) +(peak a) +(peak c) +(peak e) +(peak 3) ND +(peak f) +(peak 4) ND

+(peak 1) ND +(peak b) ND ND ND ND +(peak d) +(peak a) +(peak c) +(peak e) ND ND +(peak f) Low ND

+(peak 1) ND +(peak b) ND ND ND ND +(peak d) +(peak a) +(peak c) +(peak e) +(peak 2) ND +(peak f) +(peak 4) ND

a A “+” sign indicates that the cysteine residue was detected as free sulfhydryl. The corresponding peak numbers are in parentheses. ND indicates no detectable level of free sulfhydryl associated with cysteine residue at this location.

6456

Analytical Chemistry, Vol. 81, No. 15, August 1, 2009

marized in Table 3. It was interesting to note that there were both similarities and differences detected in the five recombinant monoclonal antibodies studied. Significant levels of free sulfhydryl were due to unpaired cysteine residues in the constant regions including CH1, CH2, and CH3 of the heavy chain and CL of the light chain. Peak b contained the first cysteine residue in the CH1 domain. Tryptic peptides containing the second cysteine residue in the CH1 domain, which was supposed to form a disulfide bond with the first cysteine residue, had a theoretical molecular weight of higher than 6 kDa. It corresponded to the small peak that eluted off the column at approximately 142 min and was confirmed by analysis of the fractions using a Q-TOF instrument with electrospray ionization (data not shown). It was not detected by MALDI-TOF most likely due to the high molecular weight and low abundance. Peak f corresponded to the peptide containing the first cysteine residue in the CL domain. The second cysteine residue in the CL domain was also found to exist in a free state to a low degree, though the levels varied for different antibodies. Variation was observed in the variable domains of the five different antibodies. Mab-C, -D, and -E shared high levels of free sulfhydryl due to the unpaired cysteine residues in their respective VH domains. The latter two, Mab-D and -E, shared the same amino acid sequences for the tryptic peptides containing the first cysteine residue in the VH domains and eluted as peak 1. Mab-C had a different amino acid sequence for the tryptic peptide with the corresponding cysteine residue and thus eluted as peak 5. A significant level of free sulfhydryl corresponding to the first cysteine residue in the VL domain was detected in Mab-A, -C, and -E but not in Mab-B and -D. These results suggested that free sulfhydryl levels in the variable domains varied significantly for the different antibodies. Interestingly, no significant levels of free sulfhydryl were detected in the hinge region. The hinge region is highly exposed, and disulfide bonds are readily accessible to solvents. The fact that no unpaired cysteine residues were detected in the hinge region further indicated that unpaired cysteine residues were formed due to the incomplete formation of disulfide bonds and not due to reduction. This result may also indicate that antibodies with unpaired cysteine residues in the hinge region were not secreted into cell culture medium or removed during purification processes. The lack of free sulfhydryl associated with cysteine residues that are involved in interchain disulfide bonds also indicated 5-IAF was efficiently removed, since interchain disulfide bonds in the hinge region are highly exposed and should be reduced preferentially by DTT and reacted with residual 5-IAF. In theory, two unpaired cysteine residues should share similar levels of free sulfhydryl if they are involved in the same disulfide bond. However, this was not always the case in the current study.

This discrepancy was likely due to several reasons. First, free sulfhydryl groups associated with two normally disulfide bonded cysteine residues may not have been equally accessible to 5-IAF. The one with a high degree of accessibility could have been modified more efficiently by 5-IAF than the one with lower accessibility. However, this cannot be due to the lack of exposure and the limit of reagent since no significant difference was observed when Mab-A was modified in the presence of 4 or 6 M guanidine hydrochloride (Figure 1) and using a 5:1 or a 10:1 5-IAF: antibody ratio (Figure 2). Second, free sulfhydryl may be chemically modified and thus not accessible to 5-IAF. For example, cysteinylation of a free sulfhydryl has been reported for a monoclonal antibody.27 Third, it was possible that low levels of free sulfhydryl groups were associated with antibodies that were not properly folded and thus contained scrambled disulfide bonds. In addition, it was worthwhile to mention that small fluorescence peaks were not collected, which most likely were peptides containing cysteine residues that were not discussed. CONCLUSIONS Recombinant monoclonal antibodies are one of the most important classes of protein therapeutics. A battery of techniques is routinely employed to characterize the structural features of these molecules. One common structural feature is the presence of low levels of free sulfhydryl. Free sulfhydryl in five recombinant monoclonal antibodies were determined after fluorescence labeling and mass spectrometry analysis. The results demonstrated that the different antibodies had different levels of free sulfhydryl. Free sulfhydryl was detected commonly in the constant domains. On the other hand, variation in the presence and levels of free sulfhydryl was observed in the variable domains. Interestingly, free sulfhydryl was rarely associated with cysteine residues that were involved in interchain disulfide bonds. The difference in the level of free sulfhydryl may relate to the difference in the production processes such as cell line, cell culture, and purification, while the similarity in the distribution of free sulfhdyrl in the domain structures may be due to the high sequence identities among the five antibodies. The use of a free sulfhydryl specific probe with a fluorescent tag facilitated detection, isolation, and identification of the locations of free sulfhydryl groups in the background of multiple disulfide bonds by HPLC and mass spectrometry. This method may be useful in the early stages of development to assist in selection of antibody candidates with higher binding affinities and better stability based on a thorough comparison of the distribution and level of free sulfhydryl groups. Received for review April 16, 2009. Accepted June 15, 2009. AC900815Z

Analytical Chemistry, Vol. 81, No. 15, August 1, 2009

6457