(CANCER RESEARCH 47, 5924-5931, November 15, 1987]
`
`New Coupling Agents for the Synthesis of Immunotoxins Containing a Hindered
`Disulfide Bond with Improved Stability in Vivo
`Philip E. Thorpe,1 Philip M. Wallace, Phillip P. Knowles, Michele G. Reif, Alex N. F. Brown, Graham J. Watson,
`Kegina E. Knyba, Edward J. Wawrzynczak, and David C. Blakey
`Drug Targeting Laboratory, Imperial Cancer Research Fund, P. O. Box 123, Lincoln's Inn Fields, London WC2A 3PX, England
`
`ABSTRACT
`
`Two new coupling agents were synthesized for making immunotoxins
`containing disulfide bonds with improved stability in vivo: sodium 5-4-
`succinimidyloxycarbonyl-a-methyl
`benzyl
`thiosulfate (SMBT) and 4-
`succinimidyloxycarbonyl-a-methyl-a(2-pyridyldithio)toluene
`(SMPT).
`Both reagents generate the same hindered disulfide linkage in which a
`methyl group and a benzene ring are attached to the carbon atom adjacent
`to the disulfide bond and protect it from attack by thiolate anions.
`An immunotoxin consisting of monoclonal anti-Thy-1.1
`antibody
`(OX7) linked by means of the SMPT reagent
`to chemically deglycosy-
`lated ricin A-chain had better stability in vivo than an immunotoxin
`prepared with 2-iminothiolane hydrochloride (2IT) which generates an
`unhindered disulfide linkage. About 48 h after i.v. injection into mice,
`one-half of the SMPT-linked immunotoxin present
`in the blood was in
`intact form and one-half as released free antibody, whereas equivalent
`breakdown of the 2IT-linked immunotoxin was seen at about 8 h after
`injection. Consequently,
`the blood levels of the SMPT-linked immuno
`toxin remained higher than those of the 211-Iinked immunotoxin despite
`loss of immunotoxin from the blood by other mechanisms. Forty-eight h
`after injection, 10% of the injected dose of the SMPT-linked immuno
`toxin remained in the bloodstream as compared with only 1.5% of the
`211-linked immunotoxin.
`The ability of immunotoxins prepared with the new reagents to inhibit
`protein synthesis by Thy-l.l-expressing
`AKR-A/2 lymphoma cells in
`vitro was identical
`to that of immunotoxins prepared with 2IT or N-
`succinimidyl-3-(2-pyridyldithio)propionate
`(SPDP). Clonogenic assays
`showed that fewer than 0.01% of AKR-A/2 cells survived exposure to
`high concentrations of OX7-abrin A-chain immunotoxins prepared with
`SMBT, 2IT, or SPDP. Twelve clones of cells which had survived
`treatment with the SMBT-linked immunotoxin were isolated. None of
`the clones was selectively resistant
`to the SMBT-linked immunotoxin
`when retested in cytotoxicity assays.
`In conclusion,
`immunotoxins prepared with the new coupling agents
`should have improved antitumor activity in vivo because they are longer
`lived and do not break down so readily to release free antibody which
`could compete for the target antigens.
`
`INTRODUCTION
`have been
`agents called "immunotoxins"
`Novel antitumor
`synthesized in several
`laboratories by covalently linking the A-
`chain of ricin and other
`toxins
`to antibodies
`against
`tumor-
`associated antigens (reviewed in Refs. 1-5). These reagents bind
`to antigens on the target cell surface, are endocytosed, and the
`A-chain then traverses the membrane, probably of the endocytic
`vesicle, and kills the cell by inactivating its ribosomes.
`Conjugation of the antibody and A-chain is generally accom
`plished by means of cross-linking agents that
`introduce a disul
`fide bond between the two proteins.
`Immunotoxins
`prepared
`with nonreducible
`linkages are consistently less cytotoxic than
`their disulfide-bonded
`counterparts
`indicating that
`reductive
`cleavage of the disulfide bond to release the A-chain in the
`cytosol may be an important
`step in the cytotoxic process (6,
`
`accepted 8/24/87.
`revised 8/21/87;
`Received 5/11/87;
`The costs of publication of this article were defrayed in part by the payment
`of page charges. This article must
`therefore be hereby marked advertisement
`in
`accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
`1To whom requests for reprints should be addressed.
`
`7). The two disulfide coupling agents used in most laboratories
`are SPDP2 (8) and 2IT (9). These agents are simple to use and
`give consistent products
`that are stable in in vitro systems and
`long-term storage.
`The disulfide bond formed by the SPDP and 2IT reagents
`appears
`to be unstable
`in vivo. We have shown that,
`after
`injection into mice, immunotoxins prepared with these reagents
`break down with a half-life of about 8 h to release free antibody.
`This is true of immunotoxins
`prepared from the A-chains of
`abrin (10), native ricin (11) and ricin which has been chemically
`deglycosylated
`to prevent
`its clearance by the carbohydrate-
`recognition
`systems of the liver (11).
`In other
`laboratories,
`evidence has been found which supports
`(12-14) or opposes
`lability (IS, 16). In those studies
`in which lability was not
`observed,
`the immunotoxins were prepared from native ricin
`A-chain and were cleared from the bloodstream so rapidly that
`the slower event of immunotoxin breakdown may not have been
`evident.
`is a problem
`Breakdown of the linkage in the immunotoxin
`for two reasons:
`(a) there is less intact
`immunotoxin
`available
`to locate and kill the tumor cells; (b) the released antibody can
`compete with the immunotoxin for the target antigens (17) and,
`being much longer-lived (11), has greater opportunity
`to bind
`to them. The effectiveness of further injections of immunotoxin
`could therefore be diminished because the tumor cell antigens
`are masked by the released antibody from the first immunotoxin
`injection.
`In the present study we synthesized two new coupling agents,
`SMBT and SMPT. These reagents were then used to prepare
`immunotoxins
`containing the same hindered disulfide linkage
`in which a methyl group and a benzene ring protect
`the disulfide
`bond from attack by thiolate anions.
`Immunotoxins
`prepared
`with the new reagents have improved in vivo stability and their
`toxicity to target cells is practically identical
`to that of immu
`notoxins prepared with SPDP or 2IT. A similar coupling agent,
`SPDB, with greater
`resistance
`to reduction was recently de
`scribed by Worrell et al. (18). In vivo stability and cytotoxicity
`data for SPDB-linked
`immunotoxins
`have not yet been re
`ported.
`
`MATERIALS AND METHODS
`
`Materials
`
`Seeds of Abrus precatorius were kindly provided by Dr. S. Olsnes
`(Norsk Hydro's
`Institute
`for Cancer Research, Oslo, Norway). The
`seeds were of Indian origin. Crushed castor beans (Ricinus communis)
`were a gift from Croda Premier Oils, Ltd., Hull, England. The beans
`were from Central Africa.
`
`2The abbreviations used are: SPDP, A'-succinimidyl-3-(2-pyridyldithio)-pro-
`pionate; SMBT, sodium S-4-succinimidyloxycarbonyl-a-methyl
`benzyl
`thiosul
`fate;
`SMPT,
`4-succinimidyloxycarbonyl-a-methyl-a(2-pyridyldithio)toluene;
`SBT, sodium S-4-succinimidyloxycarbonyl
`benzyl thiosulfate; 2IT, 2-iminothio
`lane hydrochloride; SPDB, JV-succinimidyl-3-(2 pyridyldithio)butyrate; DTT, di-
`thiothreitol;
`dg.ricA, deglycosylated ricin A-chain; abrA, abrin A-chain; 1C.,,.
`concentration
`that
`reduced [3H]leucine incorporation
`by 50%; SDS,
`sodium
`dodecyl sulfate; OX7, monoclonal antibody directed against Thy-1.1; RIO, mono
`clonal antibody directed against human glycophorin; GSH, reduced glutathione.
`
`5924
`
`Downloaded from
`
`on December 23, 2014. © 1987 American Association for Cancercancerres.aacrjournals.org
`
`
`Research.
`
`IMMUNOGEN 2088, pg. 1
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`
`HINDERED DISULFIDE COUPLING AGENTS
`
`The hybridoma cell line, MRC OX7, secreting a mouse IgGl subclass
`antibody to the Thy- 1.1 antigen, was kindly provided by Dr. A. F.
`Williams
`(MRC Cellular
`Immunology Unit, University of Oxford).
`Details of its derivation have been published by Mason and Williams
`(19). The hybridoma
`cell
`line, LICR-LON-R10,
`secreting a mouse
`IgGl subclass antibody to human glycophorin was kindly supplied by
`Dr. P. A. W. Edwards (Ludwig Institute, Sutton, England).
`The Thy- 1.1-expressing AKR-A lymphoma cell line was obtained
`from Professor!. MacLennan (Department of Experimental Pathology,
`Birmingham University, Birmingham, England).
`It was recloned to
`remove a mutant subpopulation which was resistant
`to immunotoxins
`prepared using the SPDP reagent but sensitive to immunotoxins pre
`pared using the 2IT reagent (10). The recloned line is designated AKR-
`A/2.
`Tissue culture medium RPMI 1640 and fetal calf serum were pur
`chased from Gibco-Biocult, Ltd.
`(Paisley, Scotland). Agarose (Sea
`Plaque) was from FMC Corporation (Rockland, ME). Microplates with
`96 flat-bottomed wells and tissue culture plates with 24 flat-bottomed
`wells were purchased from Flow Laboratories
`(Irvine, Scotland).
`Sodium [125I]iodide(IMS 30) and L-[4,5-3H]leucine (TRK 170) were
`obtained from Amersham International
`(Amersham, England). The
`lodo-Gen reagent for protein iodination was from Pierce (United King
`dom) Ltd. (Chester, England).
`Chromatography media were Sephacryl S-200, Sepharose 4B, Seph-
`adex G25 (fine grade), and Blue Sepharose CL-6B from Pharmacia
`Ltd. (Milton Keynes, England).
`SPDP was purchased from Pharmacia, Ltd., and 2IT from Sigma,
`Ltd. (Poole, England). Thin layer chromatography
`(SiO2) plates were
`from Merck (Kieselgel 60F). All other reagents were of analytical grade.
`
`Synthesis of SBT
`
`Bromotoluic acid (5.29 g; 25 mmol) was suspended in dioxan (10
`ml) and was mixed with a solution of sodium thiosulfate (6.45 g; 26
`mmol) in water (6 ml). The mixture was stirred at 40°Cfor 3 h during
`which time the solid dissolved and the thiosulfate derivative of toluic
`acid then crystallized out. The crystals (m.p. approximately 255°Cwith
`decomposition) were washed with cold water and dried under vacuum
`at 45°Cto constant weight (3.90 g; 14 mmol; 55%). The solid was
`dissolved in dry dimethylformamide
`(5 ml) and mixed with a solution
`of W-hydroxysuccinimide (1.82 g; 16 mmol) and dicyclohexylcarbodi-
`imide (2.97 g; 14 mmol) each in dry dimethylformamide
`(5 ml). The
`mixture was stirred for 16 h at room temperature
`and the urea was
`removed by filtration. The solvent was removed from the filtrate by
`rotary evaporation at 40 C using an oil pump and the Ar-succinimidyl
`derivative was recrystallized from methanol/CHClj.
`The yield of the
`white crystals was 3.47 g (66%). The overall yield for the synthesis was
`36%. Melting point determinations
`showed decomposition at 120'C.
`The analysis
`
`Requires: C 37.14, H 3.36, N 3.62, S 16.60, Na 5.95
`Found:
`C 36.63, H 2.78, N 3.33, S 16.27, Na 6.01
`
`was consistent with the structure shown in Fig. 1 «, ,11¡
`
`Vi i.
`
`Fig. 1. The SBT reagent (formula weight, 385.4).
`
`Synthesis
`
`of SMBT
`
`p-Ethylbenzoic acid (5.15 g; 34 mmol) was dissolved in CH2C12(45
`ml) and solid /V-bromosuccinimide (6.81 g; 38 mmol) was added fol
`lowed by benzoylperoxide (0.08 g; 0.34 mmol) in CH2C12(1 ml). The
`mixture was refluxed for 24 h. The white solid that
`remained was
`redissolved by the addition of further CH2C12 (40 ml) to the reaction
`mixture and the solution was extracted twice with water
`to remove
`succinimide. The (IMI
`solution was dried with anhydrous
`sodium
`sulfate and the solvent removed by rotary evaporation under reduced
`pressure. The white solid residue, «-bromoethylbenzoic acid, was re-
`
`to give white crystals (m.p. 140°C)
`crystallized from isopropyl alcohol
`in good yield (5.62 g; 25 mmol; 72%). A solution of a-bromoethylben-
`zoic acid (0.60 g; 2.6 mmol) in dioxan (6 ml) was mixed with a solution
`of sodium thiosulfate (0.65 g; 2.6 mmol) in water (6 ml). The mixture
`was stirred for 16 h at room temperature and the solvents were removed
`in a vacuum at 40°C.The solid was washed with (lici,
`and the
`thiosulfate derivative was recrystallized from water. The white crystals
`(shrinkage at 143"C; decomposition at 200°C)were recovered in poor
`yield (0.20 g; 0.72 mmol; 25%). The crystals were thoroughly dried and
`dissolved in dry dimethylformamide.
`The solution was mixed with
`solutions of dicyclohexylcarbodiimide
`(0.148 g; 0.72 mmol) and N-
`hydroxysuccinimide
`(0.090 g; 0.78 mmol) each in dry dimethylform-
`amide (0.4 ml). The urea that had crystallized out after
`leaving the
`solution at room temperature for 16 h was removed by filtration. The
`solvent was removed from the filtrate by rotary evaporation at room
`temperature
`using an oil pump. The oily residue was redissolved in
`methyl ethyl ketone and undissolved solid was removed by filtration.
`The solvent was removed from the filtrate by rotary evaporation under
`reduced pressure and (11(1, was added to the residue which precipi
`tated the product as a white solid that was dried under vacuum (0.17 g;
`0.43 mmol; 72%; m.p. approximately 121"C with decomposition). The
`overall yield for the synthesis was 13%. The analysis
`
`Requires: C 39.09, H 3.54, N 3.51, S 16.06, Na 5.76
`Found:
`C 38.88, H 3.68, N 3.46, S 15.69, Na 5.68
`
`was consistent with the structure shown in Fig. 2 (Ci2Hi2NOgS2Na).
`
`CH,
`
`Fig. 2. The SMBT reagent (formula weight, 399.4).
`
`Synthesis
`
`of SMPT
`
`/>-EthyIbenzoic acid was a-brominated and converted to the thiosul
`fate derivative as described in the preceding section. The thiosulfate
`(6.0 g; 20 mmol) was hydrolyzed by adding 5 N HC1 (50 ml) and stirring
`at room temperature under nitrogen for 6 h. The reaction mixture was
`then extracted three times with ethyl acetate (50 ml) and dried over
`anhydrous sodium sulfate. The solvent was then removed in a vacuum
`to leave a-thioethylbenzoic
`acid (3.0 g; 16 mmol; 80%) as a white solid
`which was stored under nitrogen.
`2-Pyridinesulfenylchloride was prepared by bubbling chlorine gas
`through a solution of 2,2-dipyridyldisuIfide
`(2.3 g; 10 mmol)
`in dry
`dichloromethane
`(20 ml) for 30 min at room temperature. The solvent
`was then removed by rotary evaporation under reduced pressure and a
`solution of a-thioethylbenzoic
`acid (1.9 g; 10 mmol) in dry dioxan (10
`ml) was added. The mixture was stirred vigorously overnight at room
`temperature
`under nitrogen. The yellow solid produced during the
`reaction was then partitioned between 0.05 M sodium phosphate buffer
`and ethyl acetate keeping the pH constant at 7.0. The organic layer
`was removed, dried over anhydrous
`sodium sulfate, and the solvent
`removed by rotary evaporation under reduced pressure to leave a yellow
`oil. Recrystallization from ethyl acetate/dichloromethane
`yielded col
`orless crystals (m.p. 130-132°C).The analysis
`
`Requires: C 57.71, H 4.50, N 4.81, S 22.01
`Found:
`C 57.92, H 4.63, N 4.89, S 21.99
`
`was consistent with the product being 4-carboxy-a-methyl-«-(2-pyri-
`dyldithio)toluene
`(C,4H13NO2S2).
`To a solution of the 2-pyridyldithio derivative (1.9 g; 6.6 mmol) in
`dry dioxan (10 ml) were added dicyclohexylcarbodiimide
`(1.4 g; 6.8
`mmol) and A'-hydroxysuccinimide (0.79 g; 6.8 mmol) each dissolved in
`dry dioxan (approximately 4 ml). The mixture was stirred for 4 h at
`room temperature,
`filtered to remove the urea, and the solvent removed
`by rotary evaporation under reduced pressure. The product was purified
`by short column chromatography on Silica Gel H. Elution was effected
`with a gradient of CH2Cl2/ethyl acetate from 0 to 50% (v/v). On
`removal of the solvent, a colorless oil (0.6 g; 1.5 mmol; 24%) remained.
`
`
`
`cancerres.aacrjournals.org Downloaded from
`
`on December 23, 2014. © 1987 American Association for Cancer
`Research.
`
`5925
`
`IMMUNOGEN 2088, pg. 2
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`
`The product was homogeneous when analyzed by thin layer chroma
`tography (Si<>•,ethyl ;tcetak-:< 11 ( I,, 1:1) but attempts to crystallize it
`were unsuccessful. The overall yield for the synthesis was 2%. Nuclear
`magnetic resonance and infra-red analyses showed ¿H(a, methyl alco
`hol) 8.36 (IH, m, pyridyl); 7.96 (2H, m, phenyl); 7.73 to 7.53 (5H, m,
`phenyl and pyridyl); 7.06 (IH, m, pyridyl); 4.30 (IH, m, CHCH3); 2.96
`(4H, S, jV-hydroxysuccinimide ester); and 1.76 (3H, m, CHCH3) *„»,
`(CH2C12) 2920, 1770, 1740, and 1605 cnT1. These analyses were
`consistent with the structure shown in Fig. 3 (C,8H|6N2Q,S2).
`
`CH3
`
`Fig. 3. The SMPT reagent (formula weight, 388.5).
`
`Purification
`
`of Abrin A-Chain
`
`Abrin was extracted from the seeds of A. precatorius by the method
`of Thorpe et al. (20). The toxin was split by reduction into its compo
`nent chains and the A-chain was purified to homogeneity as described
`previously (21). The purified toxin and the A-chain (abrA) had median
`lethal dose values of 1.4 /ig/kg and 12 mg/kg,
`respectively, when
`administered i.p. to adult BALB/c mice.
`
`Deglycosylation
`
`and Purification of Ricin A-Chain
`
`Ricin was purified from crushed castor beans by the method of
`Cumber« al. (21). A solution of the toxin (2.5 mg/ml) in 0.2 M sodium
`acetate buffer, pH 3.5, was treated for l h at 4°Cwith sodium meta-
`periodate and sodium cyanoborohydride
`at final concentrations
`of 10
`and 20 mM respectively, as described by Thorpe
`et al.
`(22). This
`procedure
`results
`in the destruction
`of approximately
`50% of the
`mannose and most of the fucose residues present on the A-chain. The
`yV-acetylglucosamine and most of the xylose residues are unaffected
`(23). The deglycosylated ricin A-chain was separated from the B-chain
`and was extensively purified by the method of Fulton et al. (24). The
`dg.ricA had a median lethal dose value of 15 mg/kg (as compared with
`30 mg/kg for native ricA) when administered
`i.p.
`to adult BALB/c
`mice.
`
`Purification of Antibodies
`
`antibodies OX7 and RIO were purified from the
`The monoclonal
`blood and ascitic fluid of hybridoma-bearing BALB/c mice by the
`method of Mason and Williams (19).
`
`Preparation of Immunotoxins
`
`Buffer Solutions. Two buffer solutions were used during the synthesis
`of the immunotoxins:
`(a) 0.05 M sodium borate, pH 9.0, containing
`1.7% (w/v) NaCI ("borate buffer");
`(b) 0.01 M Na2HPO4-0.0018 M
`KH2PO«-0.17M NaCl-0.0034 M KC1-0.001 M EDTA, pH 7.5 ("phos-
`phate-EDTA buffer").
`Derivatization of Antibody with SMBT. To a solution of antibody (20
`mg) in borate buffer (4 ml) was added SMBT (216 //I: 1 mg/ml)
`in dry
`dimethylformamide. The final concentrations
`of SMBT and antibody
`were 0.13 and 0.032 mM, giving a 4-fold M excess of SMBT over
`antibody. The solution was stirred for l h at room temperature and a
`solution of DM' (40 i/1; 15.4 mg/ml)
`in borate buffer was added, giving
`
`HINDERED DISULFIDE COUPLING AGENTS
`a molar absorptivity of 1.36 x IO4 M"1 cm"1 at 412 nm (25). The
`number of activated disulfide groups introduced using the above con
`ditions ranged between 1.5 and 1.8/molecule of antibody.
`Derivatization of Antibody with SMPT. To a solution of antibody (20
`mg) in borate buffer (2.67 ml) was added SMPT (267 Ml;0.48 mg/ml)
`in dry dimethylformamide.
`The final concentrations
`of SMPT and
`antibody were 0.11 and 0.045 mM, giving a 2.4-fold M excess of SMPT
`over antibody. The dimethylformamide was used at 10% v/v to keep
`the SMPT soluble. The solution was stirred for l h at room temperature
`and was applied to a column (30 x 1.6 cm) of Sephadex G25 (fine)
`equilibrated in nitrogen-flushed phosphate-EDTA buffer. The protein
`that eluted in the void volume of the column was concentrated to 10
`mg/ml
`in an Amicon ultrafiltration cell fitted with a YM2 membrane.
`The average number of a-methyl-a-(2-pyridyldithio)toluoyl
`groups in
`troduced into each antibody molecule was determined by reducing a
`sample of derivatized antibody solution with DTT and measuring the
`absorption of the released pyridine-2-thione which has a molar absorp
`tivity of 8.08 x IO3 M~' cnT' at 343 nm (8). The number of a-methyl-
`a(2-pyridyldithio)toluoyl
`disulfide groups introduced using the above
`conditions ranged between 1.5 and 2.0/molecule of antibody.
`Coupling of SMBT- and SMPT-derivatized Antibodies to A-chain. A
`solution of abrA (10 mg) or dg.ricA (10 mg) in phosphate-EDTA buffer
`(7 ml) was treated for 30 min at room temperature with 50 mM DTT
`and applied to a column (30 x 2.2 cm) of Sephadex G25 equilibrated
`in nitrogen-flushed
`phosphate-EDTA buffer. The A-chain fraction
`(about 35 ml) that eluted from the column was added directly to the
`concentrated antibody solution (10 mg/ml; 2 ml) in the Amicon ultra-
`filtration cell giving a molar excess of A-chain over antibody of 2.5-
`fold. The mixture was then concentrated to about 10 ml and incubated
`at room temperature for 72 h under nitrogen. The mixture was removed
`from the ultrafiltration cell and treated with 0.2 mM cysteine for 6 h at
`room temperature to inactivate any activated disulfide groups remain
`ing in the antibody component of the immunotoxin. These conditions
`do not cause splitting of immunotoxin.
`If this step were omitted, 20 to
`30% of the M, 180,000 immunotoxin interacted with plasma constitu
`ents in both in vivo and in vitro experiments
`to form a covalent adduci
`mainly of M, 240,000.
`It is possible that
`residual activated disulfide
`groups on the antibody component
`react with the thiol group of albumin
`(M, 67,000) to form the adduci.
`Preparation of SPDP- and 2IT-linked Immunotoxins. The SPDP and
`2IT coupling agents were used to link abrA or dg.ricA to OX7 antibody.
`Full details of the procedures have been published previously for both
`SPDP (6, 21) and 2IT (10, 26).
`Purification of the Immunotoxins. The products of the conjugation
`reactions above were applied to a column (90 x 2.2 cm) of Sephacryl
`S-200 equilibrated in 0.05 M sodium phosphate buffer, pH 7.5, and
`eluted with the same buffer solution. The fractions of immunotoxin
`that eluted with a molecular weight of approximately
`180,000 were
`pooled and fractionated on a Blue Sepharose column to remove free
`antibody and immunotoxin molecules containing more than one mol
`ecule of A-chain as described previously (27).
`in SDS showed that
`Analysis by polyacrylamide gel electrophoresis
`the immunotoxins
`had an apparent molecular weight of 180,000 and
`that
`they contained one molecule of antibody linked to one molecule
`of A-chain. The concentration
`of immunotoxin was determined from
`absorbance measurements
`at 280 nm. IgG-ricA (M, 180,000) has an
`1C!,1,,,at 280 nm of 1.29 assuming values of 1.40 for the antibody and
`0.765 for the A-chain (28). IgG-abrA (M, 180,000) has an E°t^at 280
`nm of 1.30 assuming values of 1.40 for the antibody and 0.787 for the
`A-chain (28).
`The immunotoxins have the structures shown in Fig. 4. The antibody
`components of the OX7 immunotoxins
`fully retained antigen-binding
`activity, as judged by fluorescence-activated
`cell sorter analyses on
`AKK A/2 cells treated with antibody or immunotoxin at saturating and
`subsaturating concentrations. The A-chain components
`fully retained
`their ability to inhibit protein synthesis in reticulocyte lysates (29).
`
`a final DTT concentration of 1 mM. The solution was stirred gently for
`a further
`l h at room temperature
`and a solution of 5,5'-dithio-bis(2-
`nitrobenzoic
`acid)
`[Ellman's
`reagent
`(25)] (40 pi; 87.2 mg/ml)
`in
`dimethylformamide was added, giving a final concentration of Ellman's
`reagent of 2.2 mM. The mixture was stirred gently for
`l h at room
`temperature
`and was applied to a column (20 x 1.6 cm) of Sephadex
`G25 (fine) equilibrated in nitrogen-flushed
`phosphate-EDTA buffer.
`The protein that eluted in the void volume of the column was concen
`trated to 10 mg/ml
`in an Amicon ultrafiltration cell fitted with a YM2
`membrane. The average number of activated disulfide groups intro
`duced into each antibody molecule was determined by reducing a sample
`Bovine IgG in borate buffer (5 mg/ml) was treated with SBT, SMBT,
`of derivatized antibody solution with DTT and measuring the absorp
`tion of the released 3-carboxylato-4-nitrothiophenolate
`ion which has
`SPDP, or 2IT to introduce an average of approximately 5 molecules of
`5926
`
`Rate Constants
`
`Downloaded from
`
`on December 23, 2014. © 1987 American Association for Cancercancerres.aacrjournals.org
`
`
`Research.
`
`IMMUNOGEN 2088, pg. 3
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`
`HINDERED DISULFIDE COUPLING AGENTS
`
`SMPT and SMBT :
`
`IgG — NH CO (o)
`
`CH3
`CH SS—
`
`A-chain
`
`SBT :
`
`SPOP :
`
`2IT :
`
`IgG — NH CO (0)
`
`CH2 SS — A-chain
`
`lgG~~ NH CO CH2 CH2 SS~~ A-chain
`
`NH2
`IgG — NH C CH2 CHj CH2 SS~~ A-chain
`
`Fig. 4. Linkages formed by the different coupling agents.
`
`coupling agent per molecule of protein. The derivatized IgG was treated
`with 1 mM DTT for l h followed by 2.2 mM Ellman's reagent for 1 h.
`This generated the same activated disulfide-leaving group (i.e., 3-car-
`boxylato-4-nitrothiophenolate
`ion) in all the derivatives. The deriva
`tized IgG preparations were desalted on columns
`(20 x 1.6 cm) of
`Sephadex G2S equilibrated
`in 0.025 M sodium phosphate,
`pH 7.4,
`containing 0.15 M NaCl.
`The IgG derivatives were treated with DTT (0.01 to 0.1 mM) or
`glutathione
`(0.1 to 1 mM) at 25°Cand the rate of release of 3-
`carboxylato-4-nitrothiophenolate
`ion was measured at 412 nm using a
`Shimadzu (Model UV 240) spectrophotometer. DTT or glutathione
`was added at
`the same time to the reference cell which contained
`underivatized bovine IgG which had been treated with DTT followed
`by Ellman's reagent in the same way as the IgG derivatives. The second
`order rate constants were calculated from the equation
`
`b(a-x)
`2.303
`t(a - b) °8a(b - x)
`
`where a is the initial molar concentration of activated disulfide groups
`in the derivatized protein solution, b is the initial molar concentration
`of DTT or glutathione, and x is the molar concentration of released 3-
`carboxylato-4-nitrothiophenolate
`ion at time l s after adding DTT or
`glutathione.
`
`Toxicity to AKR-A Cells in Tissue Culture
`
`[3H]Leucine Incorporation Assays. A suspension of AKk A/2 cells
`was prepared at 10s cells/ml
`in KI'.MI 1640 medium supplemented
`with 10% (v/v) heat-inactivated fetal calf serum, penicillin (200 units/
`ml), and streptomycin (100 /ig/ml)
`("complete medium"). The cells
`were incubated for 24 h at 37°Cwith immunotoxins
`and other
`test
`
`described previously (29).
`using the microplate method
`materials
`[3H]Leucine (1 //Ci) was then added to each culture (200 //I) and the
`radioactivity that the cells incorporated was measured 24 h later.
`Clonogenic Assays. A suspension of AKk A/2 cells was prepared at
`2 x IO5cells/ml
`in complete medium. The suspension was distributed
`in 50-ml volumes
`into 300-cm2 tissue culture flasks and complete
`medium or medium containing immunotoxin (1 ml) was added to give
`a final
`immunotoxin
`concentration
`of 1.3 x 10 * M. The cells were
`incubated for 24 h at 37°Cin an atmosphere of 5% CO2 in humidified
`air. The cells were then washed three times with complete medium.
`Cells which had been incubated in medium alone were suspended at
`120, 240, 360, 480, and 600 cells/ml
`in complete medium. Cells which
`had been treated with immunotoxin were suspended at a range of
`concentrations
`between 10' and 10 cells/ml
`in complete medium. A
`solution of 0.24% w/v agarose
`in complete medium at 45*C was
`dispensed in I ml volumes into Petri dishes (35-mm diameter) and
`cooled for
`l h at 4"( '
`to solidify the agarose. To aliquots of cell
`suspension (0.5 ml) in sterile tubes at 4"( was added a solution of
`0.24% w/v agarose in complete medium (2.5 ml) at 45°C.The suspen
`sions were then mixed, transferred in 1-ml volumes to the agar-coated
`Petri dishes, and were cooled for l h at 4°Cto solidify the agarose. The
`Petri dishes were then incubated at 37*C for 10 days and the number
`of colonies containing about 100 cells or more was counted using an
`inverted microscope. The percentage of cells that survived exposure to
`the immunotoxins was calculated by comparing their cloning efficiency
`to that of untreated cells for that particular experiment. The cloning
`efficiency of untreated cells ranged from 65 to 76%.
`
`Stability and Blood Clearance Measurements
`
`rates of the
`the stability and blood clearance
`of
`Measurements
`immunotoxins were performed as previously described (11). Briefly,
`the purified OX7-SMPT-dg.ricA and OX7-2IT-dg.ricA immunotoxins
`were radioiodinated with ' '"I to a specific activity of approximately Hi
`cpm/Mg. Groups of three adult male specific-pathogen-free BALB/c/
`ICRF mice were given injections i.v. of 10 ng of radioiodinated immu
`notoxins and samples of blood were drawn from the tail vein at various
`time intervals and transferred to heparin-coated tubes. The radioactivity
`in the blood samples (50 /<!)was measured. The samples were centri-
`fuged at 10,000 x g for 2 min and the plasma was removed. The
`radioactivity in 20 /'I plasma was counted and the samples were stored
`in liquid N2. At the end of the experiment, volumes of plasma samples
`containing 8000 cpm each were electrophoresed on 5 to 10% polyac-
`rylamide gels (1 mm thick) containing 1% SDS.
`Autoradiographs
`of the dried gels were scanned and the area under
`the immunotoxin
`(M, 180,000) peak and the released antibody (M,
`150,000) peak was divided by the total area under all
`the peaks to
`determine
`the proportion
`of radioactivity in the plasma that corre
`sponded to intact
`immunotoxin or released antibody. Calibration ex
`periments had previously shown that
`the area under each peak was
`directly proportional
`to the cpm it contained. Analysis of the immu
`notoxin by SDS-polyacrylamide
`gel electrophoresis
`under
`reducing
`conditions
`showed that
`the specific activity of the released antibody
`was somewhat
`less (9.1 x IO6 cpm/jig)
`than that of the intact
`immu
`notoxin (10 x 10' cpm/fig). Correction was therefore made for this
`difference when calculating the amount of released antibody in the
`bloodstream. Clearance measurements were expressed as a percentage
`of the injected dose assuming that the mice had a blood volume of 2.18
`ml/25 g body weight (30).
`A two-compartment
`open pharmacokinetic model was fitted to the
`plasma levels of immunotoxins
`and released antibody using a comput
`erized nonlinear
`least-squares
`regression analysis
`(31). A weighting
`function of
`\/(Y + Yf was applied to all data points
`(32). These
`analyses yielded the half-lives of the immunotoxins
`in the a and ß
`phases of clearance. Also, the half-lives of splitting of the immunotoxins
`to free antibody and A-chain were calculated using an extension of the
`same model to be described in a subsequent
`report.3
`
`RESULTS
`
`Rates of Reduction of IgG Derivatized with Various Coupling
`Agents
`
`Bovine IgG was reacted with SMBT, SBT, SPDP, or 2IT
`and then treated with DTT followed by Ellman's
`reagent
`to
`form antibody derivatives in which the same activated disulfide
`group was present
`in all. The antibody derivatives differed only
`in the groups through which the activated disulfide group was
`attached to the protein, as in Fig. 4.
`ion when
`The release of 3-carboxylato-4-nitrothiophenolate
`the antibody derivatives were treated with DTT or glutathione
`followed approximately second order kinetics, although the rate
`constant was greatest during the initial phase of reduction (Fig.
`5). The activated disulfide groups reduced first probably occu
`pied positions on the protein that were accessible to the reduc
`ing agents, whereas the more resistant groups were probably
`buried more deeply within the protein. The second order
`rate
`constants
`listed in Table 1 have been calculated at the point at
`which 50% of the leaving groups have been released and define
`the relative ease of reduction of the disulfide bonds formed by
`the different
`reagents.
`linkages depended upon the
`The stability of the different
`degree of steric hindrance
`afforded by the groups adjacent
`to
`
`3 D. C. Blakey, D. N. Skilleter, R. J. Price, H. Newell, and P. E. Thorpe,
`Comparison of the pharmacokinetics
`and hepatotoxic effects of saporin and ricin
`A-chain immunotoxins, manuscript
`in preparation.
`5927
`
`
`
`cancerres.aacrjournals.org Downloaded from
`
`on December 23, 2014. © 1987 American Association for Cancer
`Research.
`
`IMMUNOGEN 2088, pg. 4
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`
`HINDERED DISULFIDE COUPLING AGENTS
`
`100
`
`SO
`
`o —
`o. o
`5 I
`
`—
`0)
`O
`C
`ì â5
`V "-
`
`100-1
`
`75-
`
`50-
`
`25-
`
`0J
`
`eO•
`
`*-•o
`TI
`
`1013 1012io'11 io'10 109 108 io7
`
`-6
`10
`
`Concentration
`
`(M)
`
`Fig. 6. Cytotoxic effects of OX7-SMPT-dg.ricA (O) and OX7-2IT-dg.ricA (•)
`upon AKR-A/2 lymphoma cells in tissue culture. The cells were incubated for 48
`h with the immunotoxins
`or with RIO-SMPT-dg.ricA (A), dg.ricA (D), or ricin
`(•).Points, geometric means of triplicate measurements of | 'I Ijleueinc incorpo
`rated by the cells during the tinnì24-h period of culture expressed as a percentage
`of the incorporation
`in untreated cultures,
`linn, one SD about
`the mean unless
`smaller than the points as plotted. Mean [3H]leucine incorporation in untreated
`cultures was 42,000 dpm.
`
`Clonogenic assays*
`(% surviving
`
`Table 2 Cytotoxic effects ofOX7-abrA immunotoxins upon AKR-A/2
`cells in vitro
`in | '1nii-ucmc
`
`uptake assays" (M)cells)2.6
`ImmunotoxinOX7-SMBT-abrA
`x 10-
`2.0 x 10-
`3.0 x 10-
`3.3 x 10-
`>3 x 10-
`>3 x 10-
`>3 x 10-'2
`
`2
`2
`
`0.0086 ±0.0003
`OX7-SPDP-abrA
`0.0049 ±0.0021
`0.0037 ±0.0002
`OX7-2IT-abrA
`OX7-abrA cocktail'
`0.0053 ±0.0003
`RIO-SMBT-abrA
`90.0 ±8.5
`RIO-SPDP-abrA
`73.5 ±12