Bioconjugate Chem. 1990, 1, 357-362
N-Terminal Modification of Immunoglobulin Polypeptide Chains Tagged with Isothiocyanato Chelates Tariq M. Rana and Claude F. Meares' Chemistry Department, University of California, Davis, California 95616. Received July 25, 1990
Conjugatesof monoclonal antibodies with drugs, toxins, radionuclides, and other agents are in widespread use in therapeutic trials and as clinical research tools. The characterization of these immunoconjugates generally does not include determining the individual sites at which such agents are attached. We have begun to explore the attachment of the bifunctional chelating agent isothiocyanatobenzylEDTA (CITC') to the N-termini of the light chains of the Lym-1 monoclonal antibody. The similarity between this bifunctional chelating agent and Edman's reagent, phenyl isothiocyanate, led us to develop methods to distinguish between chelate-conjugated a-amino groups and t-amino groups by Edman degradation. Practically all the N-terminal Asp a-NH2 groups of Lym-1 can be modified at neutral pH, while attachment at lysine side chains predominates at pH 9. Comparison of the immunoreactivities of Lym-1-CITC conjugates with and without N-terminal conjugation shows that both are almost fully active. This implies that modification of light-chain N-termini has little or no effect on immunoreactivity, despite the fact that these residues lie near the antigen-binding sites.
Monoclonal antibody technology allows the specificity of an antibody for its antigen to be used in targeting cancer cells. Radiolabeled monoclonal antibodies (mAbs) have shown considerable promise for the early detection and therapy of cancer (1-3). As the clinical use of radiolabeled monoclonal antibodies proceeds, relatively few antibodies have been found to be useful in vivo for both diagnosis and therapy of human patients. In order to make these work better, we need improved understanding of the conjugation chemistry of immunoglobulin molecules. With the notable exception of Rodwell and co-workers ( 4 ) , most researchers in this area have used reagents that react principally with lysine residues on antibodies. In fact, vey little data is available exploring the relationship between sites of conjugation and biological activity of antibodies (other than iodination of tyrosine (5)). Since antibodies and their conjugates with drugs, toxins, radionuclides, and other useful agents are in widespread use as clinical research tools and in therapeutic trials, many of these approaches might benefit from experimenters knowing which sites on the peptide chains of the antibody can be modified successfully. The antibodies of primary interest to us are mouse immunoglobulin G (IgG) molecules, which have two pairs of polypeptide chains (light and heavy), with a total molecular weight of approximately 150K. The overall shape of an IgG molecule resembles a T or a Y (6), with the N-termini of the four polypeptide chains located near the antigenbinding sites. While the particular amino acids in the six hypervariable loops comprising the antigen-binding sites vary with antigen specificity, most of the residues in the light and heavy chains do not change from one antibody to another. Thus, the chemical properties of one IgG should provide a useful guide to others. The reactive groups we consider here are NH2 groups 1 Abbreviations: t-Boc, tert-butoxycarbonyl; CAPS, 3-(cyclohexy1amino)-l-propanesulfonicacid; CITC, isothiocyanatobenzylEDTA; DNFB, 2,4-dinitrofluorobenzene;DNP, dinitrophenyl; mAb, monoclonal antibody; MCA, 7-amino-4-methylcoumarin;
PTH, phenylthiohydantoin. 1043-1802/90/ 2907-0357$02.50/0
located at lysine side chains and at the polypeptide chain termini (the a-amino groups of polypeptide chains react with the same reagents as lysine c-amino groups). Mouse IgG2a mAb Lym-1 contains 30 lysines in each of ita y2a heavy chains and 13 lysines in each of its KVlight chains, for a total of 86 lysines in the molecule (7). The lysines are distributed throughout the light and heavy chains of the antibody. In mouse immunoglobulins 95% of the expressed light chains are of the K type, and 80% of these contain Asp as the N-terminal residue (8, 9). The vaat majority of subgroup KVlight chains (95% ) have Asp aa their N-terminal residue (9). Light chains of Lym-1 also showed Asp as their N-terminal residue during sequence analysis (Table 11). Heavy chains of Lym-1 have Gln as the N-terminal residue and showed no sequence during automated Edman sequence analyses. The primary structural analyses of many naturally occurring proteins and peptides including gastrin, fibrinopeptides, collagen, and immunoglobulins have been hindered by the presence of pyrrolidonecarboxylic acid (pyroglutamic acid) as an amino terminal residue (10, and references therein). It is believed that cyclization of terminal glutaminyl or glutamyl reaidues leads to the formation of pyroglutamic acid. Sanger and Thompson (11)observed the spontaneous formation of pyroglutamyl peptides during the isolation and proteolysis of peptides containing glutamine as N-terminal residues. Since the automated Edman sequence analysis depends on the availability of an a-amino group, the proteins beginning with pyroglutamic acid pose problems in their direct sequencing. Because the NH2 termini of the Lym-1 heavy chains are blocked, the only a-aminogroups available for conjugation are on the light chains. The usual strategy with monoclonal antibodies is to seek labeling conditions that will not block the antigenbinding sites. Because an IgG antibody molecule contains so many lysine residues, and because the great majority of these are not found in the antigen-binding sites (82/ 86 for Lym-1),lysine t-amino groups are usually considered good targets. 0 1990 American Chemical Society
358
Rana and Meares
Bioconjupte Chem., Vol. 1, No. 5, 1990
Scheme I I I I I
I
I
Coupling
I
I / / /
/ / /
/
J
/ /
Cleavage
/ / /
,,
/
0
0
,,
,,
0
1. PTH-amino acids (R = H) 2. CITC-Gly (R CHzEDTA, R1 H)
,,
0
3. CITC-ASP(R = CHzEDTA, R1 = CH$OOH)
To a first approximation, it might be expected that chemistry directed toward a particular residue such as lysine would produce a statistical mixture of products (12). For example, if all lysines were equally reactive, then modifying one of the 86 lysines in Lym-1 would involve a lysine in an antigen-binding site with a probability of only 4/86, or 4.6%. Assuming that any hit in an antigenbinding site inactivates the antibody, but any other hit has no effect, this would produce a conjugate with about 95 % of its original immunoreactivity. We have frequently observed results contrary to that simple expectation. Conjugation yields and immunoreactivity of conjugates vary with reaction conditions in complex ways that depend strongly on the pH of the reaction, among other factors (13, 14). Since the PKa of a-amino groups is 7.6-8, while the pKa of €-aminogroups is 9.3-9.5, and since it is the unprotonated RNH2 groups that react with electrophilic reagents (15), one possibility to be investigated is whether conjugation of the a-amino groups with bifunctional chelating agents affects antigen binding. This is particularly interesting, since these a-amino groups lie within 10 A of the antigen-binding sites (16). Typical conjugation reactions in our laboratory involve addition of isothiocyanatobenzyl-EDTA (CITC) to an antibody (17). The similarity between this bifunctional chelating agent and Edman’s reagent, phenyl isothiocyanate (Scheme I), suggested to us that we could distinguish between chelate-conjugated a-amino groups and e-amino groups by Edman degradation. The literature contains numerous examples of the use of analogues of phenyl isothiocyanate bearing fluorescent tags for protein microsequencing (e.g. 18-22). Here, we report experiments demonstrating how a chelate isothiocyanate can reveal N-terminal conjugation and some of the properties of such conjugates.
-
EXPERIMENTAL PROCEDURES
Chemicals. Diglycine, triglycine, and Asp-Ala peptides were purchased from Sigma Chemical Co. The peptide t-Boc-Gly-Lys-Arg-MCAwas purchased from Peptides International. Bromophenol blue and 2,4-dinitrofluorobenzene (DNFB) were purchased from Aldrich. Lym1,an anti B cell lymphoma IgGza mAb (23), was obtained from Damon Biotech (Needham Heights, MA; Encapcel murine mAb, lot # 3-171-860813). It was further purified
LA
P I
1
1
12
L I
I
18
25
Retention time (min) Figure 1. HPLC profile of the products of Edman degradation of CITC-triglycine, after treatment with DNFB. Retention times of authentic compounds are indicated: DNP-diglycine, 12 min; DNFB, 18 min. The peak at 17 min is presumably the CITCGly Edman product. The shape of the solvent gradient is indicated by the dashed line; see Experimental Procedures for details.
by protein A affinity column chromatography prior to use. Protein A on Sepharose-CL-4B, NaDodSOd, Tris, CAPS, and Coomassie Blue R-250 were obtained from Sigma Chemical Co. PVDF membranes (Immobilon transfer), 0.45 pm pore size, were obtained from Millipore. Cobalt57 chloride was purchased from ICN (specific activity 7000 Ci/g). Pure water (resistance 18 Ma, NANOpure 11, Barnstead, MA) was used throughout the experiments. All glass labware was washed with a mixed acid solution and thoroughly rinsed with pure water (24). All plastic labware was washed with 3 M HC1 and thoroughly rinsed. All other chemicals were the purest grade available. Thin-Layer Chromatography. TLC was run on plastic-backed silica gel plates (EM Science) using a solution composed of equal volumes of 10% (w/v) aqueous ammonium acetate and methanol as the eluent. In this system, protein conjugates remain at the origin while free chelates and smaller peptide-chelate conjugates migrate to Rf 0.4-0.8. High-Performance Liquid Chromatography. Reversed-phase HPLC for analyses of CITC-peptide conjugates, DNP-diglycine, DNP-triglycine, and reaction mixtures after Edman degradation was performed at room column (Alltech). temperature with a 10 X 250 mm A 20-min linear gradient, from 0.1 M ammonium acetate, pH 7.0 (containing 1mM EDTA), to 100% methanol, was used for analyses of CITC conjugates. A gradient system shown in Figure 1,from 0.1 M ammonium acetate, pH 7.0 (0.1 mM EDTA), to 90% acetonitrile, was used to isolate DNP derivatives, at a flow rate of 3.0 mL/min. The UVabsorbing fractions were detected at 254 nm. Purification of peptide-CITC conjugates, DNPdiglycine, and DNP-triglycine was done by reversedphase HPLC using a 21.4 X 250 mm c18 column (Dynamax). Gradients and solvent systems were the same as described above, with a flow rate of 12.5 mL/min. Radiation Counting. y-Counting was done in a Beck-
N-Terminal Attachment
man Model 310 counter with the energy window set for 57C0. TLC plates containing radiolabeled materials were visualized with an AMBIS radioanalytical imaging system. Spectroscopy. Proton NMR spectra were recorded on a QE 300 spectrometer at 300 MHz. IR spectra were recorded on an IBM IR/32 spectrometer. Exact mass measurements were obtained by running low- and highresolution mass spectra on a ZAB-HS-2F mass spectrometer (VG Analytical, Wythenshawe, UK). During mass spectroscopic measurements, either 3-nitrobenzyl alcohol or dithiothreito1:dithioerythritol (3:l w/w) was used as a matrix along with small amounts of p-toluenesulfonic acid. High-resolution FAB spectra contained polyethyleneglycol or polyethyleneglycol methyl ether as reference compound. Preparation of the Peptide-Conjugates. DNPdiglycine and DNP-triglycine were prepared according to Sanger's method (25). Each peptide (0.4 g) and 0.8 g of NaHC03 were dissolved in 10 mL of water, and a solution of 0.8 g of DNFB in 10 mL of ethanol was added to this mixture. T h e mixture was shaken for 3 h a t room temperature, concentrated under reduced pressure to remove the ethanol, dissolved in water, and extracted with ether to remove the excess DNFB. The aqueous solution was acidified, causing the separation of an oil that immediately solidified in amorphous form. DNPpeptides were purified by HPLC, and characterization was done by proton NMR and TLC. Isothiocyanatobenzyl-EDTA (CITC) was prepared as described by Meares et al. (17). Triglycine (0.5 mmol) was dissolved in 0.1 M sodium phosphate, pH 8.0, to give a 0.1 M triglycine solution; 0.5 mmol of CITC was dissolved in the same buffer to the same final concentration. Conjugation was started by mixing both solutions and adjusting the final pH to 9.0. The reaction was carried out with constant stirring at 40 "C for 18 h. The course of the reaction was monitored by fluorescamine test (26) for the free amine terminus of triglycine. The reaction product gave one spot on TLC with R f 0.6. The product was purified by HPLC, and characterized by proton NMR, FTIR (absence of SCN stretch at 2100 cm-l), and FABMS ( m / e 629, M + 1). t-BOC-Gly-Lys-Arg-MCA peptide (8.5 pmol) was dissolved in 0.1 M sodium phosphate, pH 8.0, to give a 20 mM peptide solution; CITC was dissolved in the same buffer to the same final concentration. CITC was attached to the e-amino group of the single lysine residue in the peptide by mixing the two solutions and adjusting the final pH to 9.0. The reaction was carried out with constant stirring at pH 9.0 and 40 OC for 5 h. The course of the reaction was monitored by fluorescamine test. The product was purified by HPLC and confirmed by FAB-MS ( m / e 1056, M + 1). Asp-Ala peptide (24.5 pmol) was dissolved in 0.1 M sodium phosphate, pH 8.0, to give a 20 mM peptide solution; CITC was dissolved in the same buffer to the same final concentration. The conjugate of CITC and AspAla was prepared by mixing the two solutions and adjusting the final pH to 9.0 with saturated Na3P04 (aqueous).The reaction was carried out with constant stirring at pH 9.0 and 40 "C for 10 h. The course of the reaction was monitored by fluorescamine test for the free amino terminus of aspartate. The product was purified by HPLC and confirmed by FAB-MS ( m / e 644, M + 1). Edman Degradation of Peptide-CITC Conjugates. This was carried out by the method of Chang (27) with the following modifications. CITC-triglycine conjugate was dried under reduced pressure and 1-5 pmol was dissolved in 500 pL of anhydrous trifluoroacetic acid in
Bloconjugate Chem., Vol. 1, No. 5, 1990 359 Table I. Lym-1-CITC Conjugation Reactionsa reaction p H chelates/Lym-1 % immunoreactivitvb
7.0 9.0
2.6 2.9
* *
91.6 3.6 96.3 4.8
*
Reaction conditions described in Experimental Procedures. Average values SD from triplicate immunoreactivityassays. (I
*
an Eppendorf tube; the tube was flushed with Nz, capped, and heated at 54 "C for 15 min. The trifluoroacetic acid was evaporated, 200 pL of water was added, and the pH was adjusted to 9.0. After the cleavage reaction, diglycine was converted to DNP-diglycine by adding an excess of DNFB to the reaction mixture. This conversion was carried out by stirring the reaction mixture a t room temperature and pH 9.0 for 3 h. Excess DNFB was extracted with ether, and the aqueous layer was analyzed by HPLC. Edman degradation of the t-Boc-Gly-Lys(C1TC)-ArgMCA conjugate was done as described above. The products were analyzed by FAB-MS. Edman degradation of the CITC-Asp-Ala conjugate was performed as described above. After the cleavage reaction, cleaved CITC-Asp was converted to PTH-like derivative 3 (shown in Scheme I) by adding 25 % aqueous TFA and heating at 54 "C for 5 min, evaporated to dryness, and analyzed by FAB-MS and Applied Biosystems 470A gasphase sequencer. Preparation of Lym-1-CITC Conjugates. Lym-lCITC conjugates were prepared at different pH conditions. The number of chelates per antibody was determined by 57C0assay (17). The Lym-1 antibody solution (15-20 mg/ mL) was prepared for conjugation with a centrifuged gelfiltration column (28) with 0.1 M sodium phosphate, pH 8.0 or 6.0, as the column buffer. The conjugation reactions were carried out as described below. Excess CITC was removed, and conjugates were transferred to 0.1 M ammonium citrate, pH 6.0, by centrifuged gel-filtration column chromatography. The course of the conjugation reactions was followed by subjecting aliquots of the reaction mixtures to 57C0assay. pH 9.0 Conjugate. CITC was dissolved in 0.1 M sodium phosphate buffer, pH 8.0, and added to Lym-1 antibody solution in the same buffer (final concentrations: Lym1 , O . l mM; CITC, 0.6 mM). The pH of the solution was adjusted to 9.0 and the reaction mixture was incubated at 37 OC for 3 h. pH 7.0 Conjugate. CITC solution in 0.1 M sodium phosphate buffer, pH 8.0, and Lym-1 antibody solution in the same buffer were mixed together (final concentrations: Lym-1,O.l mM; CITC, 25 mM), and the final pH was adjusted to 7.0. The reaction mixture was incubated at 37 "C for 2 h. Immunoreactivity Assay. Solid-phase radioimmunoassays for immunoreactivity of either lllIn- or 57C0labeled conjugates were done in triplicate as reported previously (29)with 1251-labeledantibody as the standard. Immunoreactivity values given in Table I are relative to 1251-labeledantibody. NaDodS04-Polyacrylamide Gel Electrophoresis. The protein gels used in these experiments employed the NaDodSO4 system of Laemmli (30). The running gel was 20% acrylamide with a thickness of 0.5 mm and a length of 7 cm. Gels were run in a Bio-Rad Mini-PROTEAN I1 system for 1 h at 80 V and then at 200 V until the bromophenol blue tracking dye had run off the bottom of the gel (ca. 1h). After electrophoresis was complete, the gels were rinsed in electroblotting transfer buffer (see below).
960
Rana and Meares
Bioconlugate Chem., Vol. 1, No. 5, 1990
Table 11. Comparison of N-Terminal Sequences of Two Lym-1-CITC Conjugates pH 9.P cycle AA pmol AA pmol 1 ASD 150 "TrD" "30" 2 Ile120 Ile 42 3 Gln 100 Gln 31 4 Met 70 Met 30 5 Thr 120 Thr 37 33 6 Gln 90 Gln I Ser 50 8 Pro 70
pH 7.0
AA Ile Gln Met Thr Gln Ser
pmol 16 10 13 22 18 40
The pH 9.0 conjugate shows the normal sequence.
Electroblotting. Transfer of proteins from polyacrylamide gels to PVDF membranes was carried out according to the procedure of Matsudaira (31). The membrane was cut to the same size as the gel, wetted with a brief rinse in 100% methanol, and then equilibrated by soaking in transfer buffer for 10-15 min. The transfer buffer employed was 10 mM CAPS, 10% methanol, pH 11.0, to reduce the level of Tris and glycine in the medium. The gel, sandwiched between a sheet of PVDF membrane and several sheets of blotting paper, was assembled into a BioRad Trans Blot apparatus and electroeluted for 30 min at 10 "C and 0.5 A in transfer buffer. A backing sheet of nitrocellulose was included to catch any protein that passed through the first membrane. The PVDF membrane was washed in deionized water for 5 min, stained with 0.1 % Coomassie Blue R-250 in 50% methanol for 5 min, and then destained in 50% methanol, 10% acetic acid for 10 min at room temperature. The membrane was finally rinsed in deionized water for 10 min, air-dried, and stored at -20 "C. Edman Degradation of Lym-1-CITC Conjugates, Sequencing of Lym-1 and Lym-1-CITC conjugates was performed by automated Edman degradation with subsequent identification of the PTH-amino acids by chromatographic methods. This was done on an Applied Biosystems 470A gas-phase sequencer. Data reduction was achieved with Nelson Analytical software on an IBM A T computer. RESULTS AND DISCUSSION
Isothiocyanatobenzyl-EDTA (CITC) can cleave N-terminal amino acids in the same way as phenyl isothiocyanate, the Edman degradation reagent. This was confirmed by labeling the N-terminus of triglycine with CITC. The N-terminal labeled triglycine-CITC conjugate was subjected to Edman degradation conditions in anhydrous trifluoroacetic acid at 54 "C for different time periods, and the cleavage products were analyzed by TLC. The N-terminal labeled triglycine-CITC conjugate gave one UV-quenching spot on TLC with R f 0.6 which was radioactive (when CITC was labeled with 57C0),but fluorescamine negative, which shows the absence of free amino groups. After 5 min of trifluoroacetic acid treatment two spots appeared on the TLC, one UV-quenching, and fluorescamine-negative spot with the same Rf (0.8) as CITC (which was radioactive when Wo-CITC was used) and the other a nonquenching, nonradioactive, fluorescaminepositive spot at the Rf (0.5) of diglycine. In this TLC system, triglycine gave one fluorescamine-positivespot at RfO.4. After trifluoroacetic acid treatment for 10 min, the cleavage was complete. Extended treatment with trifluoroacetic acid for 15 min did not affect the intensity of the two product spots. The experimental outline for Edman chemistry on the peptide-CITC conjugates and the structures of the
Scheme I1
1
CFsCOOH
products are shown in Scheme I. After cleavage of N-terminal glycine, diglycine was converted into DNPdiglycine as described in Experimental Procedures. DNP derivatives of triglycine and diglycine were synthesized, and their retention times on Cla reversed-phase HPLC were identified. HPLC analysis showed that one of the cleavage products had the same retention time as DNP-diglycine (Figure 1). These results indicate that the N-terminal glycine residue was removed from CITC-triglycine during trifluoroacetic acid treatment. The chelate CITC does not cleave amino acids from peptides when attached to the t-amino group of lysine. This was confirmed by a control experiment performed by conjugation of the CITC to a lysine-containing peptide with a blocked N-terminus. The experimentalsteps are outlined in Scheme 11. A conjugate of CITC and t-Boc-Gly-LysArg-MCA was synthesized and treated with trifluoroacetic acid under anhydrous conditions at 54 "C for 15 min. The reaction products were dried and analyzed by FABMS ( m / e 957, M + 1). The results showed that the t-Boc group was lost from the peptide (as expected), but CITC was still attached. Lym-1-CITC conjugates were prepared at pH 7 and 9 as described in Experimental Procedures. Heavy and light chains of Lym-1 were isolated by NaDodSOr gel electrophoresis and electroblotted onto Immobilon membranes for sequencing. The heavy chain of unmodified Lym-1 has a blocked N-terminus, therefore sequencing of Lym-lCITC conjugates was carried out for the light chain only. The light chain of the Lym-1-CITC conjugate prepared at pH 9, which had 2.9 chelates/mAb, showed the normal sequence of the light chain through eight Edman cycles: Asp-Ile-Gln-Met-Thr-Gln-Ser-Pro (Table 11). On the other hand, the light chain of the Lym-lCITC conjugate prepared at pH 7 , with a chelate/mAb ratio of 2.6, showed no Asp at all in the first Edman cycle.
N-Terminal Attachment
As shown in Table 11, a mixture of two amino acids was observed: one eluted near Trp and the other was Ile (a “preview” of the actual second residue in the chain). The second Edman cycle showed a mixture of the expected Ile and a preview of the next residue Gln, and so on. These data indicate that practically all the N-terminal Asp a-NH2 groups were modified during the pH 7 conjugation. Further, they indicate t h a t t h e resulting adduct is somewhat unstable, since a significant number of the light chains analyzed had lost their original N-terminal residue. Finally, they show that the CITC-Asp Edman product elutes near T r p during sequence analysis. Further identification and characterization of the CITCAsp Edman degradation product was done by synthesizing a CITC-Asp-Ala conjugate. After HPLC purification, the CITC-Asp- Ala conjugate was subjected t o Edman degradation and the UV-absorbing product was isolated and analyzed by FAB-MS ( m / e555, M + 1). These results confirm that CITC-Asp was cleaved from the CITC-AspAla conjugate during trifluoroacetic acid treatment and cyclized to a PTH-like derivative 3 (shown in Scheme I). The CITC-Asp Edman product eluted near T r p on the protein sequencer, as did the first Edman degradation cycle product of the Lym-1-CITC conjugate prepared at pH 7.0. The instability of the N-terminal CITC-Asp adduct is worthy of comment. The pH 7 conjugate was not exposed to strongly acidic conditions prior to Edman analysis. However, Asp residues in peptides are well-known to promote hydrolysis of adjacent peptide bonds under mildly acidic conditions (32-35), presumably through t h e interaction of the un-ionized side chain COOH of Asp with the amide carbonyl oxygen. This internal acid might promote Edman degradation of the N-terminal CITCAsp adduct under very mild conditions, providing a possible explanation. As shown in Table 11, N-terminal analysis of the immunoconjugate readily reveals the presence or absence of the CITC-Asp adduct. To the extent that it would lead to loss of radiolabel from a target-bound antibody, such instability is not desirable for most in vivo applications. However, separation of chelate from antibody might be advantageously accelerated in the (pH 5) intracellular compartments of the liver. Further investigation is required to determine which effect is more important. Comparison of the immunoreactivities of the pH 7 and 9 Lym-1-CITC conjugates is also instructive. Respectively, they were 91 % and 96% of control 1251-Lym-1(Table I). Since practically all of the light-chain N-termini were modified in the pH 7 conjugate, but practically none in the pH 9 conjugate, this implies that modification of lightchain N-termini has little or no effect on immunoreactivity, despite the fact that these residues lie near the antigen-binding sites. Further experiments with the preferential labeling of these N-termini with other reagents could lead to stable immunoconjugates retaining all the antibody effector functions found in the constant regions that lie far away from the N-termini. ACKNOWLEDGMENT We thank Daniel Jones for running the FAB-MS, Gary Mirick for performing the immunoassays, John Gardner for sequencing the proteins, Michael McCall for the preparation of CITC and technical advice on numerous aspects of these experiments, and Sally and Gerald DeNardo for helpful discussions. Supported by NIH Grant Pol-CA47829 (G. L. DeNardo, P. I.).
Bioconjugate Chem., Vol. 1, No. 5, 1990 301
LITERATURE CITED (1) Cope, D. A,, Dewhirst, M. W., Friedman, H. S., Bigner, D.
D., and Zalutsky, M. R. (1990) Enhanced delivery of a monoclonal antibody F(ab’)2 fragment to subcutaneous human glioma xenografts using local hyperthermia. Cancer Res. 50, 1803-1809. (2) Williams, J. A., Wessels, B. W., Edwards, J. A., Kopher, K. A., Wanek, P. M., Wharam, M. D., Order, S. E., and Klein, J. L. (1990)Targeting and therapy of human glioma xenografts in vivo utilizing radiolabeled antibodies. Cancer Res. 50 (3 Suppl.), 974s-979s. (3) Fritzberg, A. R., Berninger, R. W., Hadley, S. W., and Wester, D. W. (1988) Approaches to radiolabeling of antibodies for diagnosis and therapy of cancer. Pharm. Res. 5, 325-334. (4) Rodwell, J. D., Alvarez, V. L., Lee, C., Lopes, A. D., Goers, J. W. F., King, H. D., Powsner, H. J., and McKearn,T. J. (1986) Site-specific covalent modification of monoclonal antibodies: in vitro and in vitro evaluations. Proc. Natl. Acad. Sei. U.S.A. 83,2632-2636. (5) Roholt, 0.A., and Pressman, D. (1967)Differential method for determining the relative reactivity to iodination of different tyrosyl residues in a protein molecule. Biochim. Biophys. Acta 147,l-14. (6) Silverton, E. W., Navia, M. A., and Davies,D. R. (1977)Threedimensional structure of an intact human immunoglobulin. Proc. Natl. Acad. Sei. U.S.A. 74, 5140-5144. (7) Wellman, A. A., and Meares, C. F. (1990) Sequences of the Lym-1 Antibody Heavy and Light Chain Variable Regions. Nucl. Acids Res. 18, Issue 17. EMBL Accession Numbers X53483 and X53484. (8) Zachau, H. G. (1989) Immunoglobulin Light-Chain Genes of the K Type in Man and Mouse. Immunoglobulin Genes (T. Honjo, F. W. Alt, and T. H. Rabbits, Eds.) pp 91-109, Academic Press, London. (9) Kabat, E. A,, Wu, T. T., Reid-Miller,M., Perry, H. M., and Gottesman, K. S., (Eds.) (1987) Sequences of Proteins of Immunological Interest, 4th ed, Public Health Service, National Institutes of Health, Bethesda, MD. (10) Podell, D. N.,and Abraham, G. N. (1978)A Technique for the Removal of Pyroglutamic Acid from the Amino Terminus of Proteins Using Calf Liver Pyroglutamate Amino Peptidase. Biochem. Biophys. Res. Commun. 81, 176-185. (11) Sanger, F., and Thompson, E. 0. P. (1953) The Aminoacid Sequence in the Glycyl Chain of Insulin. Biochem. J . 53, 366-374. (12) Chang, C-H., Meares, C. F., and Goodwin, D. A. (1982) Bifunctional Chelating Agents: Linking Radiometals to Biological Molecules. In Applications of Nuclear and Radiochemistry (Lambrecht,R. M., and Morcos, C. N., Eds.) pp 103-114, Pergamon, New York. (13) Mirzadeh, S.,Brechbiel, M. W., Atcher, R. W., and Gansow, 0. A. (1990) Radiometal Labeling of Immunoproteins: Covalent Linkage of 2-(4-Isothiocyanatobenzyl)diethylenetriaminepentaaceticAcid Ligands to Immunoglobulin. Bioconjugate Chem. 1, 59-65. (14) McCall, M. J., Diril, H., Meares, C. F. (1990) Simplified Method for Conjugating Macrocyclic Bifunctional Chelating Agents to Antibodies via 2-iminothiolane. Bioconjugate Chem. 1, 222-226. (15) Means, Gary E., and Feeney, Robert E. (1971) Chemical Modification of Proteins, pp 11-17, Holden-Day, Inc., San Francisco, CA. (16) Segal, D. M., Padlan, E. A., Cohen, G. H., Rudikoff, S., Potter, M., and Davies, D. R. (1974)The Three-Dimensional Structure of Phosphorylcholine-Binding Mouse Immunoglobulin Fab and the Nature of the Antigen Binding Site. Proc. Natl. Acad. Sei. U.S.A. 71, 4298-4302. (17) Meares, C. F., McCall, M. J., Reardan, D. T., Goodwin, D. A,, Diamanti, C. I., and McTigue, M. (1984) Conjugation of Antibodies with Bifunctional Chelating Agents: Isothiocyanate and Bromoacetamide Reagents, Methods of Analysis, and Subsequent Addition of Metal Ions. Anal. Biochem. 142, 68-78. (18) Chang, J. Y. (1988) A complete quantitative N-terminal analysis method. Anal. Biochem. 170, 542-556.
362 Bioconjugate Chem., Vol. 1, No. 5, 1990
(19) Vergeres, G., Winterhalter, K. H., and Richter, C. (1989) Identification of the membrane anchor of microsomal rat liver cytochrome P-450. Biochemistry 28, 3650-3655. (20) Tsugita, A. (1987) Developments in protein microsequencing. Adv. Biophys. 23, 81-113. (21) Amoscato, A. A., Babcock, G. F., Sramkoski, R. M., Hynd, B. A., and Alexander, J. W. (1987) Synthesis of two biologically active fluorescent probes of thymopentin. Int. J . Pep. Protein Res. 29, 177-186. (22) Miyano, H., Nakajima, T., and Imai, K. (1987) Microscale sequence analysis from the N-terminus of peptides using the fluorogenic Edman reagent 4-N,N-dimethylamino-lnaphthyl isothiocyanate. Biomed. Chromatog. 2, 139-144. (23) Epstein, A. L., Zimmer, A. M., and Spies, S. M. (1985) Radioimmunodetection of human B-cell lymphomas with a radiolabeled tumor-specific monoclonal antibody (Lym-1). In Malignant Lymphomas and Hodgkin’s Disease:Experimental and Therapeutic Advances (F. Cavalli, G. Bonadonna, and M. Rozencweig, Eds.) pp 569-577, Martinus Nijhoff Publishing Co., Boston. (24) Thiers, R. C. (1957) Contamination in Trace Element Analysis and Its Control. Methods Biochem. Anal. 5, 273335. (25) Sanger, F. (1945) The Free Amino Groups of Insulin. Biochem. J . 39,507-515. (26) Udenfriend, S., Stern, S., Bohlen, P., Dairman, W., Leimgruber, W., and Weigele, M. (1972) Fluorescamine: a reagent for assay of amino acids, peptides, proteins, and primary amino acids in the picomole range. Science (Washington,D.C.) 178, 871-872. (27) Chang, J-Y. (1983) Manual Micro-Sequence Analysis of Polypeptides Using Dimethylaminoazobenzene Isothiocyanate. Method Enzymol. 91,455-466.
Rana and Meares (28) Penefsky, H. S. (1979) A Centrifuged-Column Procedure for the Measurement of Ligand Binding by Beef Heart F1. Methods Enzymol. 56, Part G , 527-530. (29) DeNardo, S. J., Peng, J.-S. B., DeNardo, G. L., Mills, S. L., and Epstein, A. L. (1986) Immunochemical Aspects of Monoclonal Antibodies Important for Radiopharmaceutical Development. Nucl. Med. Biol. 13, 303-310. (30) Laemmli, U. K. (1970) Cleavage of Structural Proteins During the Assembly of the Head of Bacteriophage T4. Nature 227,680-685. (31) Matsudaira, P. (1987) Sequence from Picomole Quantities of Proteins Electroblotted onto Polyvinylidene Difluoride Membranes. J. Biol. Chem. 262, 10035-10038. (32) Fraser, K. J., Poulsen, K., and Haber, E. (1972) Specific cleavage between variable and constant domains of rabbit antibody light chains by dilute acid hydrolysis. Biochemistry 11,4974-4977. (33) Hanford, R., Doonan, H. J., Doonan, S., Vernon, C. A., Walker, J., Bossa, F., Barra, D., Carloni, M., Fasella, P., and Riva, F. (1974) Primary structure of aspartate aminotransferase from pig heart muscle. Peptides produced by cleavage with cyanogen bromide and with dilute acid. Atti Accad. Naz. Lincei, C1. Sci. Fis., Mat. Nat., Rend. 56, 73-83. (34) Jauregui-Adell, J., and Marti, J. (1975) Acidic cleavage of the aspartyl-proline bond and the limitations of the reaction. Anal. Biochem. 69,468-473. (35) Inglis, A,, McKern, N., Roxburgh, C., and Strike, P. (1980) Cleavage of proteins a t aspartic acid residues using dilute acid. Methods of Peptide and Protein Sequence Anulysis. Proceedings of the 3rd International Conference;Meeting Date 1979 (Birr, Christian, Ed.) pp 329-343. Elsevier Amsterdam, Netherlands.
