2 Antibodies for Delivering Saporin to Lymphoma in Vitro and in Vivo1

[CANCER RESEARCH 51. 2353-2361, May 1, 1991] Cooperative Mixtures of Bispecific F(ab')2 Antibodies for Delivering Saporin to Lymphoma in Vitro and in Vivo1

Ruth R. French, Ann E. Courtenay, Susan Ingamells, George T. Stevenson, and Martin J. Glennie2

Lymphoma Research Unit, Tenovus Laboratory, General Hospital, Southampton, United Kingdom

ABSTRACT in which one Fab' was specific for the surface idiotype of the guinea pig lymphoblastic leukemia, L2C, and the second Fab' We report that selected combinations of two or more monoclonal bispecific F(ab')i antibodies (BsAbs) far outperform single derivatives in recognized a plant RIP, saporin (7, 8). During this work, a simple and efficient means was developed for producing large the delivery of the ribosome-inactivating protein, saporin, to guinea pig amounts of pure BsAb, in which individual Fab' arms were I.;( leukemic cells. Throughout the work, BsAbs were constructed by thioether-linking the hinges of two Fab'7, one from monoclonal anti- joined at their hinges via a stable thioether linkage. The only L2C-idiotype antibody and the other from anti-saporin antibody. The prerequisite for preparing these hybrid molecules was that the latter was either affinity-purified rabbit poh donai or one of a panel of parent antibodies, polyclonal or monoclonal, yielded F(ab')2 five mouse monoclonal antibodies. In vitro cytotoxicity studies showed fragments on enzymatic digestion that could be reduced to Fab' that, though all derivatives were effective, the BsAb made with the fragments expressing free hinge-region sulfhydryl groups for poh donai antibody was always 10 to 20 times more potent than those cross-linking by o-phenylenedimaleimide. BsAb proved highly made with a monoclonal antibody in yielding 50% inhibition of [3H|- efficient at delivering saporin to the L2C cells in vitro and leucine uptake. This superior activity could be matched by selective prevented tumor growth in vivo as efficiently as conventional mixtures of two or more of the monoclonal derivatives. Furthermore, in saporin-containing immunotoxin (8). immunotherapeutic delivery of saporin to tumor, a pair of BsAbs per Throughout this work, BsAb derivatives constructed using formed significantly better than did either individually. Fab' fragments from an affinity-purified rabbit polyclonal anti- Binding and uptake studies with radiolabeled saporin demonstrated a 20-fold increase in functional affinity when saporin was held at the cell saporin antibody performed more efficiently than did similar surface by an appropriate BsAb mixture rather than by a single BsAb. reagents containing monoclonal anti-saporin antibodies. The In contrast, only small differences were recorded in the rate at which specificity for the L2C cells has always been supplied by Fab' saporin was internalized as a result of the same maneuver. We conclude from a high-affinity monoclonal anti-idiotype antibody (9). The that the improved performance of combinations of BsAbs arises from advantage of the polyclonal antibody-containing derivative was their ability to provide multiple linkages between saporin molecules and most evident in immunotherapy experiments, where it was cell surfaces, significantly increasing the functional affinity with which approximately 10 times more effective than a similar derivative saporin is tethered to the cell, but, in this system at least, having only a containing monoclonal anti-saporin antibody. In this work, we minor effect on the rate at which it is internalized. Cocktails of two or more BsAbs, selected to bind to multiple epitopes on ribosome-inactivat have used a panel of five monoclonal anti-saporin antibodies to ing proteins and perhaps also on unwanted cells, could provide an prepare BsAb derivatives. In vitro and in vivo studies show that, important new strategy in immunotherapy. although none of the individual monoclonal bispecific deriva tives matches the polyclonal reagent, a cocktail of all five, or selected pairs, of the monoclonal derivatives performed equally INTRODUCTION as well as did the original polyclonal reagent. The mechanism Conventional immunotoxins, prepared by covalently cou of this improved performance is investigated. pling plant or bacterial toxins to appropriate antibodies, have been shown to be selectively toxic to unwanted cells both in MATERIALS AND METHODS vitro and in vivo (1, 2). As an alternative approach to targeting Materials. Cell culture was performed in supplemented DMEM cytotoxic compounds, Raso and Griffin (3) and Raso (4) dem containing glutamine (200 ITIM),pyruvate (100 HIM), penicillin and onstrated the potential of bispecific antibodies with dual speci streptomycin (100 lU/ml), Fungizone (2 mg/ml), and 10% PCS (My- ficity for target cells and the RIP3 ricin as a means of delivering oclone; Gibco, Ltd., Paisley, Scotland). The RIP saporin was purified toxic substances to neoplastic cells. They showed that BsAbs from the seeds of Saponaria officinalis by water extraction as described having specificity for ricin A chain and the surface immuno- previously (10). globulin of a lymphoma line could deliver toxin to the target Leukemia Cells. The L2C lymphoblastic leukemia cells of strain 2 cells and prevent further protein synthesis. Other workers have guinea pigs were used throughout this investigation (11, 12). The line is maintained by continuous passage in vivo; animals receiving an developed and modified this concept to include different types inoculum of IO5cells die 13 to 15 days later with a WBC count close of BsAbs for targeting various pharmacological agents, includ to 250,000/jil. Cells were recovered by flotation on Ficoll-Hypaque ing Vinca alkaloids (5, 6). More recently, we have constructed BsAb F(ab')2 derivatives (Lymphoprep; Nycomed as., Oslo, Norway) as described previously (12). Cell viability was always >90% as judged by trypan blue exclusion. Received 10/1/90; accepted 2/19/91. Antibodies and Antibody Fragments. For this study, five mouse mono The costs of publication of this article were defrayed in part by the payment clonal antibodies, anti-sap-1 to -5, were raised against saporin using of page charges. This article must therefore be hereby marked advertisement in the method of KÃ¶hlerand Milstein (13) as described by Fazekas de St. accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1This work has been supported by Tenovus of Cardiff, the Cancer Research Groth and Scheidegger (14). Hybridoma cells secreting anti-saporin Campaign, and Italfarmaco, Milan, Italy. antibody were identified by ELISA and cloned by limiting dilution in 2To whom requests for reprints should be addressed. microculture plates. The other monoclonal antibody used to construct 'The abbreviations used are: RIP, ribosome-inactivating protein; BsAb, bi the BsAb was an anti-idiotype antibody, anti-Id-1, which is a mouse specific antibody; PBS, phosphate-buffered saline: BSA. bovine serum albumin: IgGl specific for the IgM idiotype of the L2C lymphoblastic leukemia ICso, 50% inhibitory concentration; K,, association (equilibrium) constant; DMEM, Dulbecco's modified Eagle's medium; PCS, fetal calf serum; ELISA, (9). enzyme-linked immunosorbent assay; IgGl, immunoglobulin Gl; IgM, immu- All hybridoma lines secreting these antibodies were expanded as noglobulin M: SH, sulfhydryl. ascitic tumors in Pristane-primed BALB/c x CBA F, mice. 2353

Polyclonal antisera reacting with saporin and guinea pig l;;il>V (15) been precoated overnight with 7S IgG isolated from rabbit anti-saporin were raised in rabbits and rats, respectively, using standard immuniza antibody (20 Mg/ml in sodium carbonate buffer, pH 9.6). Test samples tion protocols (7). and antibody standards were then added, and antibody bound to the The 7S IgG fractions of ascitic fluid and polyclonal antiserum were saporin was detected with the appropriate dilution of horseradish isolated by precipitation in 2 M ammonium sulfate, followed by ion peroxidase-labeled rabbit anti-mouse immunoglobulin (Nordic Labo exchange chromatography on Trisacryl-M-DEAE (9). F(ab')2 frag ratories, Ltd., Maidenhead, Berkshire, United Kingdom). Each stage ments from both monoclonal and polyclonal IgG were prepared by of the assay consisted of a 90-min incubation at 37Â°C,followed by five limited proteolysis with pepsin at pH 4.2 as described previously (7, 9). washes in PBS to remove unbound reagent. Knowing the starting Preparation of BsAb. Heterodimeric F(ab')2 molecules containing concentration of antibody, we could then calculate the amount of two different mouse Fab' fragments or a mixture of mouse and rabbit antibody bound to saporin in the equilibrium mixture and plot the Fab' fragments were constructed as described by Glennie el al. (7). results by the method of Scatchard (18) to determine the K,s. For these Briefly, F(ab')2 molecules from the required antibody were reduced to calculations, it was assumed that each IgG molecule had the capacity obtain Fab' fragments with hinge region SH groups, Fab'(SH). The to bind two molecules of saporin. SH groups on one of the Fab'(SH) species were then fully alkylated In the second procedure, the binding of 125I-saporin to immobilized with excess o-phenylenedimaleimide to provide free maleimide groups. monoclonal antibody was determined. Flexible 96-well plates (Titertek; The two preparations, Fab'(mal) and Fab'(SH), were then combined Flow Laboratories, Ltd., Rickmansworth, United Kingdom) were under conditions that allowed cross-linking of the maleimide and the coated with anti-saporin monoclonal IgG (10 Mg/ml in sodium carbon SH groups and avoided reoxidation of the SH groups. The final ate buffer, pH 9.6) overnight at 4Â°Cand then blocked with PBS products were reduced and alkylated to remove any minor untoward containing 1% BSA for l h at 37Â°C.After washing 3 times with PBS, products that may have formed by oxidation or disulfide exchange, and !>l saporii) was added at various concentrations, and the plates were the final mixture was fractionated according to size by chromatography incubated for a further l h at 37Â°C.Following final washing, the plates on Ultragel AcA44 (LKB-Produkter AB, Bromma, Sweden). were air dried, and each well was cut out for radioactive counting in a Incorporation of |3H|Leucine by Cultured L2C Cells. The incorpora Rackgamma spectrometer (LKB). K,s were determined by plotting the tion of (3H]leucine into protein during short-term culture of L2C cells results according to the method of Scatchard (18). has been described previously (7, 8). Briefly, complexes of BsAb and Binding of I25l-Saporin to Cell Surfaces in the Presence of BsAb. The saporin were preformed for l h and then incubated with fresh L2C cells binding of '"I-saporin to the cell surfaces in the presence of BsAb was (5 x 105/well) for 7 h at 37Â°C,before pulsing overnight with 0.5 ^Ci of investigated using a method based on that described by Dower et al. pH]leucine (TRK.510, Amersham International, Amersham, United (19). Free and cell-bound antibody in an equilibrium mixture can be Kingdom). The incorporation of |'H]leucine into cell protein was then separated by rapid sedimentation of the cells through a mixture of assessed by harvesting the cells onto glass microfiber filters and washing phthalate oils, leaving the aqueous phase as a discrete layer above the with water. All experimental points on the graph were determined in oil. In these experiments the BsAb was preincubated with I25l-saporin triplicate. The concentration of saporin at which [3H]leucine uptake by to allow complexes to be formed before the addition of L2C cells. Any cells was inhibited by 50% was taken as the IC50value. radioactivity that sedimented with the cell pellet was therefore due to To determine the kinetics of protein synthesis inhibition, 100-^1 '"I-saporin linked to the cell surface via BsAb. samples of BsAb and saporin at the appropriate concentration in Radiolabeled saporin was serially diluted and incubated as 1-ml supplemented leucine-free RPMI 6400 (Gibco) were incubated for l h aliquots with BsAb at 1 Mg/ml in supplemented DMEM at 37Â°Cfor 1 at 37Â°Cin 96-well microculture plates before chilling to 4Â°C.Fresh h. A 100-^1 volume of fresh L2C cells (5 x 107/ml) was then added, L2C cells (5 x 105/well), which had been preincubated for 2 h in leucine- and the incubation continued for a further l h at 37Â°C.The concentra free medium at 37Â°Cand chilled, were then added to each well for a tions of BsAb and the concentration range of 125I-saporinin this mixture further 2 h at 4Â°C.Microculture plates were then transferred to 37Â°C were kept the same as those used in the in vitro [3H]leucine uptake in a humidified atmosphere of 5% CO2 in air and, at the required time experiments described above. Any endocytosis of saporin/BsAb com points, wells were pulsed with 1 Â»iCiof['H]leucine in 50 M'of supple plexes was prevented by inclusion of sodium azide (15 mM) and 2- mented leucine-free RPMI 6400 for 30 min. Incorporation of [3H]- deoxyglucose (50 mM). The cells, with any bound saporin/BsAb com leucine was stopped by the addition of 30 M' of a mixture of 5 mivi plexes, were separated from the aqueous phase by centrifugation cycloheximide and 20 mg/ml of L-leucine in PBS. At the end of the through phthalate oils as described by Elliot et al. (9). experiment, the incorporation of ['Hjleucine into cell protein was Uptake and Internalizaron of Targeted I25l-Saporin by Cultured M assessed by harvesting as described above. Each time point was deter Cells. The extent to which radiolabeled saporin, once bound to L2C mined in triplicate, and the results were expressed as a percentage of cells by single or mixtures of monoclonal BsAbs, was internalized was the incorporated counts obtained in cells incubated for the same period followed during short-term culture. To distinguish between the total in medium alone. 125I-saporin associated with the cells and that residing in internal Radioiodination of Proteins. Saporin was trace radiolabeled for bind compartments, we used mild proteolysis with the enzyme papain to ing studies using carrier-free '"I (Radiochemical Centre, Amersham, strip material from the cell surfaces. Previous work from this laboratory United Kingdom) and IODO-BEADS (Pierce Chemical Co., Rockford, has shown that guinea pig cell surface IgM, to which the BsAb and IL) as the oxidizing reagent (16). Radioactivity was measured in a consequently the '25I-saporin would be bound, is highly sensitive to Rackgamma spectrometer (LKB). proteolysis by papain (12). Antibody Binding Constants. Two methods were used to determine BsAb and saporin were first preincubated at the appropriate concen the binding affinity of each of the five anti-saporin monoclonal anti tration in 20 ml of supplemented DMEM for l h at 37Â°Cto allow bodies. In the first procedure, the affinity of the antigen-antibody complexes to form. Following chilling to 4Â°C,IO8fresh L2C cells were binding in free solution was measured using an ELISA procedure as added in 2 ml of cold supplemented DMEM, and the mixture was described by Friguet et al. (17). First, 50 ng/ml (anti-sap-1) or 100 ng/ incubated for a further 2 h to permit the BsAb/saporin complexes to ml (anti-sap-2 to -5) of monoclonal IgG in PBS containing 1% BSA equilibrate at the cell surface. The temperature was then raised to 37Â°C were incubated with various concentrations of saporin overnight at 4Â°C. in a water bath, and at various times samples taken and stored on ice The concentration range of saporin appropriate for each monoclonal in the presence of 15 mM sodium azide to prevent further metabolic antibody was determined in preliminary experiments: 2 to 125 Mg/ml activity. for anti-sap-2 and anti-sap-4 and 0.2 to 25 Mg/ml for anti-sap-1, anti- To determine intracellular 125I-saporin, duplicate 1-ml aliquots of sap-3, and anti-sap-5. The concentration of the free IgG antibody cells were washed once by centrifugation (230 x g for 5 min) in DMEM remaining in the equilibrium mixture was then measured using the to remove excess FCS. Following resuspension in 1 ml of DMEM, the following indirect ELISA procedure. Saporin (0.2 Mg/ml in PBS con cells were exposed to preactivated papain at 0.2 mg/ml for 30 min in taining 1% BSA) was captured on 96-well microculture plates that had the presence of micrococcal endonuclease (5 ng/ml) as described pre- 2354

Downloaded from cancerres.aacrjournals.org on September 23, 2021. © 1991 American Association for Cancer Research. BISPECIFIC ANTIBODIES FOR THE DELIVERY OF SAPORIN TO LYMPHOMA viously (12). Finally the cells were centrifugea 3 times in PBS contain ing 1% BSA and 10 HIMsodium azide to remove material, including surface-bound 125I-saporin, stripped from the cells by papain treatment, and the cell pellets were counted in a Rackgamma spectrometer (LKB). The total 125I-saporin associated with the cells was determined by rapid centrifugaron of the cells through the mixture of phthalate oils 80 â€¢ as described above, and the cell pellets were counted for radioactivity. Immunotherapy Studies. Groups of ten age-matched strain 2 guinea pigs (420 to 480 g) were inoculated i.p. with 10s fresh L2C cells (Day 0) in 1 ml of DMEM. At 24 h (Day 1), animals were treated i.p. with a mixture of BsAb and saporiti in 1 ml of PBS at the concentrations described in the experiments. Surviving animals were recorded daily, and results were analyzed by the x2 test of Peto (20). 8 20 â€¢

C o RESULTS ra o Delivery of Saporin to L2C Cells in Culture. BsAbs were fi. constructed to link saporin to the idiotype of surface IgM on L2C cells. All derivatives used Fab' fragments from the same Â¡ high-affinity (K* 1.1 x IO9 M"') monoclonal anti-idiotype anti o> 'o body, anti-Id-1, to provide their target cell specificity. In con 3 trast, the anti-saporin binding specificity was acquired from Fab' prepared from one of five monoclonal anti-saporin anti bodies, anti-sap-1 to -5, or from affinity-purified rabbit IgG anti-saporin. The efficiency of these derivatives at delivering saporin to the L2C cell to cause inhibition of protein synthesis was determined first in vitro. Uptake of [3H]leucine was meas ured in cells cultured in medium containing a constant concen tration of the appropriate BsAb derivative (1 M8/ml), together with various concentrations of saporin (8). As shown in Fig. \A, in the absence of a BsAb, saporin was comparatively non- 10 â€¢10 10 -9 10 â€¢ 10 - 10 -6 toxic, giving half-maximal uptake of [3H]leucine (IC50) at ap Saporin (M) proximately 3.3 x 10~7 M (10 Mg/ml). However, when an Fig. 1. Uptake of [3H]leucine by LÂ¡Ccells cultured in medium containing saporin and BsAb. Cells (5 x IO5) were cultured in supplemented DMEM appropriate BsAb was included, the IC50 values dropped dra containing saporin at the concentrations shown and BsAb at 1 Mg/ml for 8 h at matically. In the case of the five monoclonal anti-saporin- 37Â°C.[3H]Leucine (0.5 pCi) was added and the incubation continued for a further containing derivatives (anti-sap-1 to -5 x anti-Id-1), IC50values 12 h. The cells were then harvested, and the incorporation of radioactivity was varied between 3 x 10~9M and 2.5 x 10~'Â°M(Table 1). Despite determined. A, saporin alone (x); anti-sap-1 x anti-Id-1 (â€¢);anti-sap-2 x ami Id 1 (O); anti-sap-3 x anti-Id-1 (â€¢);anti-sap-4 x anti-Id-1 (A); anti-sap-5 x ami-Id minor variations between experiments, the performance of 1 (D); cocktail of all five BsAbs (total concentration, I >jg/ml) (T); rabbit anti-sap x anti-Id-1 (A). B, anti-sap-1 x anti-Id-1 (â€¢);ami-sap-1 x anti-Id-1 + anti-sap-3 these BsAbs relative to each other remained the same: anti-sap- x anti-Id-1 (total concentration, 1 /Â¿g/ml)(â€¢);anti-sap-1 x anti-Id-1 + anti-sap- 1 > anti-sap-5 > anti-sap-3 > anti-sap-4 > anti-sap-2. 4 x anti-Id-t (A); ami sap 1 x anti-Id-1 + anti-sap-5 x anti-Id-1 (D); cocktail of It is also clear from Fig. \A that none of the monoclonal all five BsAbs (T). anti-saporin-containing derivatives was as efficient as the polyclonal reagent (rabbit anti-sap x anti-Id-1) in this assay. Its IC50 of 1.5 x 10~" M was at least 16 times lower than that of the best of the five monoclonal anti-saporin derivatives (anti- Table 1 Performance of single or mixtures of BsAbs al delivering saporin to sap-1 x anti-Id-1). Interestingly, a cocktail containing equal cultured L2C cells quantities of the five monoclonal BsAbs matched (Fig. \A; of saporin/cell Table 1) or, in some experiments, slightly improved on the ExperimentBsAbA atIC50*3,5003,50011,5003,5007,0004,0005,5004,0005.000of performance of the rabbit anti-sap x anti-Id-1. This was inves NoneAnti-sap-1 tigated further to determine the minimum number of reagents IdAnti-sap-2X ami anti-Id-Anti-sap-3x which, when mixed, might provide a similar level of perform anti-Id-Anti-sap-4x ance. Fig. IB and Table 1 (Experiment B) show that, with amildAnti-sap-5x appropriate selection, only two monoclonal anti-saporin-con ami-IdRabbitx taining derivatives were necessary to attain a "minimum mix anti-Id-1Cocktailanti-sap x ture" whose performance was close to that of the full cocktail BsAbsB of 5 or the polyclonal reagent. From a number of experiments we Anti-sap-1anti-Id-1Cocktail x BsAbsAnti-sap-1of 5 established that the derivative anti-sap-1 x anti-Id-1, when +anti-sap-3x anti-Id-1 mixed with an equal amount of anti-sap-3 x anti-Id-1, anti-sap- anti-Id-1Anti-sap-1x 4 X anti-Id-1, or anti-sap-5 x anti-Id-1, could deliver saporin +anti-sap-4x anti-Id-1 anti-Id-1Anti-sap-1x as effectively as the full cocktail. Adding a third monoclonal +anti-sap-5x anti-Id-1 anti-saporin derivative to any of these pairs neither improved anti-Id-1Â° x maximalobtainedSaporin concentrations giving ':â€¢ nor reduced their activity. In addition, none of the other pairs proteinsynthesiscell 1.* from experiments shown in Fig. that could be formulated from the five monoclonal reagents moleculesobtainedThe average number of saporin perMoleculesoat the 1CÂ»was performed any better than did anti-sap-1 x anti-Id-1 alone. from the experiments shown in Fig. 4.10'Â°XICS3,3002.53012206.50.150.352.60.50.320.20.55inhibitionbound 2355

100 librium binding constants obtained from Scatchard plots for the five monoclonal antibodies are given in Table 2. Represent ative Scatchard plots for one higher (anti-sap-1) and one lower (anti-sap-2) affinity anti-saporin antibody are shown in Fig. 3. The saturation binding curves are also included in Fig. 3, insets. Comparable results were obtained from the two assay proce dures, with Kt values in the range of 5 x IO6 M~' to 5 x IO7 M"1. Anti-sap-4 failed to give a clear saturation curve, so we were unable to calculate a precise association constant. For anti-sap-1, the affinity was determined for both the parent anti- saporin and its bispecific derivative, anti-sap-1 x anti-Id-1. The similarity of their A'.,values indicates clearly that the procedures 18 20 22 24 26 28 30 32 Days after inoculation used to construct BsAb do not perturb antigen binding in any Fig. 2. Immunotherapy with BsAb. Groups of ten age-matched guinea pigs untoward way. were inoculated i.p. with 10' fresh 1 ..( cells on Day 0. After 24 h (Day 1), animals were treated i.p. with saporin (10 UK)and BsAb (SO Â»ni(molar ratio, 1.5:1): the control group received PBS alone. Survival was recorded daily. Control (â€¢):mui sap-1 x anti-Id-1 (O); anti-sap-3 x anti-Id-1 (x); anti-sap-1 x anti-Id-1 + anti- sap-3 x anti-Id-1 (â€¢).

Table 2 K^for the binding of monoclonal antibody to saporin K.S were calculated from Scatchard plots (see Fig. 3). Results are representative of at least 3 similar experiments. (M-') by (M-') by AntibodyAnti-sap-1 ELISA5.2 I25l-Saporinbinding4.1 x IO7 x IO7 Anti-sap-2 4.5 x IO6 9.0 x 10' Anti-sap-3 5.2 x IO7 2.8 x IO7 Anti-sap-4 <4.5 x 10' <9.0 x 10' Anti-sap-5 3.5 x IO7K. 1.4 x IO7 Anti-sap-1 x anti-Id-1K. 5.2 x IO7

Immunotherapy. Immunotherapy studies were performed in L2C tumor-bearing guinea pigs to compare the performance of 2.0 2.5 0 0.5 B (moles saporin/mole Ig) the individual BsAb with that of pairs of BsAbs. Saporin (10 vg) and BsAb at a 1:1.5 molar ratio were premixed and injected i.p. into guinea pigs that had received an inoculum of 10s I..-C' 50 cells 24 h previously. This is a suboptimal dose of therapy in the L2C tumor model (8), which was chosen to allow small 12 Â¡40 increases in animal survival to be detected. We have shown previously that, when highly effective therapeutic regimens are adopted (e.g., 25 to 100 Â¿Â¿gsaporin/animal),almost all animals 8.30 survive beyond 26 days and then succumb to an idiotype- negative variant of the L2C tumor. Once a "maximum thera peutic effect" has been achieved in this way, it is not possible 20 to detect any minor difference in the performance of derivatives. The results of the suboptimal immunotherapy are shown in Fig. 2. By treating with a combination of anti-sap-1 x anti-Id- 2 10 1 and anti-sap-3 x anti-Id-1, clear evidence was obtained that a minimum mixture was more therapeutic than was the individ ual BsAb (pair versus anti-sap-1 x anti-Id-1, P < 0.01; pair 02468 10 05 10 15 20 25 versus anti-Sap-3 x anti-Id-1, P = 0.08). Furthermore, this 10" x B (moles saporin/well) mixture was also more effective than the cocktail of all five Fig. 3. Representative Scatchard plots for the binding of anti-sap monoclonal antibody to saporin assessed by two assay methods. Plots for the binding of anti- monoclonal anti-saporin-containing derivatives (P < 0.03; re sap-1 (A and ( ') and anti-sap-2 (B and />) are shown. In .1 and B (Assay 1), a sults not shown). constant amount of anti-sap IgG (50 or 100 ng/ml) was incubated with various Binding Affinity of the Monoclonal Anti-Saporin Antibodies. concentrations of saporin overnight at 4*C. The concentration of free antibody remaining was then determined by an ELISA procedure with reference to a We next investigated the relation between the binding affinity standard curve for the appropriate antibody. The results are plotted as bound of the five parent monoclonal anti-saporin antibodies and the saporin/free saporin [mol of saporin mol of IgG/M (HT)] versus bound saporin (It). Insets, saturation binding curves. In C and D (Assay 2), 96-well plates were performance of the corresponding BsAbs. Antibody binding coated with anti-sap IgG overnight and then incubated with various concentra affinity was estimated by two methods: (a) using an ELISA to tions of '"I-saporin for l h at 37*C. The plates were then washed, and the wells determine the affinity of antibody-antigen interaction in free were cut out and counted to determine the bound radioactivity. Knowing the specific activity of the >25I-saporin, the amount bound/well was calculated. The solution and (/>) by radioimmunoassay to measure the affinity results are plotted as bound saporin/free saporin [mol/well/M (B/F)] versus bound of radiolabeled saporin binding to immobilized antibody. Equi saporin (B). insets, saturation binding curves. 2356

Downloaded from cancerres.aacrjournals.org on September 23, 2021. © 1991 American Association for Cancer Research. BISPECIFIC ANTIBODIES FOR THE DELIVERY OF SAPORIN TO LYMPHOMA Binding of I25l-Saporin to L2C Cells in the Presence of BsAb. but not saporin. However, by far the most striking uptake of To determine if combinations of BsAb bound saporin to cells saporin was achieved with the cocktail of all five reagents, more avidly than did single reagents, studies were performed where saturation was reached at much lower concentrations of to measure the binding of radioiodinated saporin to L2C in the saporin than with any of the single BsAbs (Fig. 4A, inset). In presence of various derivatives. In these experiments, cell- the presence of the cocktail, half maximal binding of I25I- bound and free '25I-saporin were separated by rapid centrifu- saporin to the cell was achieved at 5 x 10"'Â°Mcompared with gation of the cells through a mixture of phthalate oils. The >1 x 10~" M for anti-sap-1 x anti-Id-1, anti-sap-3 x anti-Id-1, binding of 125I-saporin to L2C cells was determined under the and anti-sap-5 x anti-Id-1. This difference in binding represents same conditions (saporin concentration range and antibody an increase in functional affinity of at least 20-fold. Interest concentration) as those used in the toxicity assays (Fig. 1). Fig. ingly, although the cocktail bound saporin more avidly than 4A shows the binding curves obtained in the presence of single any of the individual reagents, the maximum number of I25I- BsAb and the cocktail of all five derivatives. The full 125I-saporin saporin molecules bound per cell at saturation was significantly binding curves are shown in the inset, and the binding levels less. For example, while anti-sap-1 x anti-Id-1 bound at least obtained with low concentrations of 125I-saporin are shown on 1.5 x 10*saporin molecules/cell, the cocktail of reagents bound an expanded scale in the main figure. As expected, the most only 9 x IO4molecules/cell. significant 125I-saporin binding was achieved with those BsAbs containing the higher affinity anti-saporin antibodies, anti-sap- Similar binding studies were performed to compare the per 1, -3, and -5. BsAbs containing the lower affinity antibodies, formance of minimal mixtures of two BsAbs. Fig. 4B shows that those minimal mixtures that had performed well in the anti-sap-2 and -4, gave saporin binding which was only slightly toxicity assay and in immunotherapy bound I25I-saporin to the higher than with a control BsAb with specificity for the cells target cells as avidly as did the cocktail of all five reagents, half maximal binding of saporin being reached around 5 x 10~'Â°M.

80 r Furthermore, as with the cocktail, at saturation fewer molecules of saporin were bound to each cell than with the individual BsAb. Using these binding data, we were able to estimate, for each 60 BsAb and each antibody mixture, the number of saporin mole cules bound to target cells at the saporin IC50 values obtained in toxicity studies (Fig. 1). The results in Table 1 show that, in each case except one, half maximal inhibition was achieved when between 3,500 and 7,000 molecules of saporin were bound to each cell. The one exception to these findings was the ffl O derivative anti-sap-3 x anti-Id-1 which needed to bind approx imately 11,000 molecules of saporin to each cell before reaching QQ. half-maximal inhibition. CO Kinetics of Inhibition of | '1111cucine Uptake. The rate at which M 10 -9 saporin inhibited [3H]leucine uptake into cells in the presence _ Ã¼ 80 of BsAb was examined. Chilled L2C cells were incubated with V 125I-saporin and BsAb to obtain surface binding, the tempera O ture was then increased to 37Â°C,and the uptake of [3H]leucine B X was determined at various times. For each derivative or com bination of derivatives, a concentration of saporin was selected that was just sufficient to achieve maximal inhibition of [3H]- leucine uptake in the toxicity assays described above (Fig. 1).

40 The 125I-saporin binding studies had shown that these condi tions were fulfilled when approximately 30,000 molecules of saporin were bound to each cell (Fig. 4). This meant that we could measure the rate at which a constant level of saporin was transported into the cells to switch off protein synthesis. The results in Fig. 5 demonstrate that all the BsAbs caused rapid inhibition of protein synthesis in cultured L2C cells with no 10" 10 indications of any lag phase while the saporin took effect. Saporin (M) Individual monoclonal BsAbs did show some differences in Fig. 4. Binding of I2'l-saporin lo L2C cells in the presence of BsAb. BsAb (1 Â»ig/ml)wasincubated with various concentrations of 125I-saporin for l h at 37'C. their kinetic profiles: anti-sap-1 x anti-Id-1 and anti-sap-2 x L2C cells (5 X 10") were then added, and the incubation continued for a further anti-Id-1 required less than 3 h to inhibit [3H]leucine uptake by 1 h. Any endocytosis of cell-bound complexes was inhibited by including IS IHM 90%, while anti-sap-3 x anti-Id-1 and anti-sap-5 x anti-Id-1 sodium azide and 50 mM 2-deoxyglucose (9). The cells were then sedimented through phthalate oils, and the pellet was counted for radioactivity. The results needed over 5 h to achieve this level of inhibition. In contrast, are expressed as the number of molecules of saporin bound/cell. Insets, full the performance of the cocktail of all five BsAbs and the saturation curves. The main figures show the binding obtained at the lower minimal mixtures gave kinetic profiles that were almost iden concentrations of saporin on a log scale after subtracting the binding obtained with an irrelevant BsAb without specificity for saporin..). anti-sap-1 x anti-Id-1 tical to each other and very similar to those of anti-sap-1 x (â€¢):anti-sap-2 x anti-Id-1 (O); anti-sap-3 X anti-Id-1 (X); anti-sap-4 x anti-Id-1 anti-Id-1 and anti-sap-2 x anti-Id-1 when used individually. As (â€¢);anti-sap-5 x anti-Id-1 (A); cocktail of BsAbs (O); irrelevant BsAb (A). B, anti-sap-1 x anti-Id-1 + anti-sap-3 x anti-Id-1 (x); ami sap-1 x anti-Id-I + anti- expected under these conditions, saporin in the absence of a sap-5 x anti-Id-1 (A); cocktail of BsAbs (O). BsAb did not inhibit the uptake of | 'II jleueinc. 2357

Downloaded from cancerres.aacrjournals.org on September 23, 2021. © 1991 American Association for Cancer Research. BISPECIFIC ANTIBODIES FOR THE DELIVERY OF SAPORIN TO LYMPHOMA course, as shown by an increase in the total cell-associated I25I- saporin and in the I25l-saporin in intracellular compartments and so protected from papain digestion. The initial rate of uptake of saporin in the cells with the individual BsAbs and with the mixed reagents was approximately 200 and 300 molecules/cell/min, respectively. By 4 h, 125I-saporin accumulation had reached a plateau indicating that uptake of fresh 125I- saporin/BsAb complexes was almost matched by loss of '"!- saporin from the cells. For the individual BsAbs, anti-sap-1 x anti-Id-1, anti-sap-3 x anti-Id-1, and anti-sap-5 x anti-Id-1, the levels of saporin that accumulated per cell over the 4-h time course were around 15,000, 21,000, and 22,000 molecules, respectively. Slightly higher levels (21,000 to 29,000 molecules/ cell) were achieved with the mixtures of BsAb (Fig. 6, D and E) and also with the cocktail of all five derivatives (Fig. 6F). We next measured 125I-saporin uptake into cells when com plete internalization of surface-bound l25I-saporin/BsAb com plexes was induced by a modulating antibody. It was anticipated that this experiment would give some measure of the maximal uptake of l25I-saporin/BsAb complexes which could be achieved in this system and an indication of how close uptake in the presence of BsAb came to this goal. It was important for this work that the antibody used to induce modulation cross-linked the surface IgM directly and did not disturb the binding equi librium between the saporin/BsAb complexes and the surface IgM; the antibody chosen, rat anti-guinea pig Fabju, did not Time (h) bind mouse Fab' or saporin and was known to induce almost Fig. 5. The kinetics of saporin toxicity in the presence of BsAb. L:C cells (5 x 10*) preincubated in leucine-free medium were incubated with BsAb (1 pg/ml) complete loss of surface IgM from L2C cells by antigenic and saporin at the required concentration (see below) for 2 h at 4*C and then transferred to 37*C. At the times shown, wells were pulsed with 1 >iCi of [3Hj- modulation (21). A single BsAb, anti-sap-3 x anti-Id-1, was allowed to equili leucine for 30 min, and the incorporation was stopped by the addition of 30 M!of 5 mM cycloheximide -t-20 mg/ml of L-leucine. At the end of the time course, the brate with I25l-saporin at the cell surface, and then rat anti- cells were harvested, and the incorporation of radioactivity was determined. Saporin alone, 1 ng/ml (x); anti-sap-1 x anti-Id-1 + 0.1 Mg/ml of saporin (â€¢); guinea pig Fatyi was added to induce modulation. The results anli-sap-2 x anti-Id-1 + 1 jig/ml of saporin (O); anti-sap-3 x anti-Id-1 and anti- in Fig. 6(7 show a pattern of I25l-saporin uptake that is quite sap-5 x anti-Id-1 + 0.25 fig/ml of saporin â€¢and A. respectively); anti-sap-1 x distinct from that induced by any single or mixture of BsAb: anti-Id-1 + anti-sap-3 x anti-Id-1 + 0.025 Mg/ml of saporin (D); anti-sap-1 x (a) the level of total cell-associated I25l-saporin (surface plus anti-Id-1 + anti-sap-5 x anti-Id-1 + 0.025 Mg/ml of saporin (A); cocktail of five BsAbs -t-0.025 Mg/ml of saporin (V). intracellular) was unchanged over the 4-h time course; and (b) there was a sharp drop in the level of surface-bound saporin over the first 1 h that corresponds exactly to the material Uptake of I25l-Saporin into L2C Cells in the Presence of BsAb. accumulated in the cells over the same period (30,000 mole We next investigated the uptake of 125I-saporin by L2C cells in cules/cell). As a result of the modulating anti-Fab^ antibody, the initial rate of I25l-saporin uptake was increased from 200 the presence of single or combinations of BsAbs. Chilled L2C cells were incubated with 125I-saporin and BsAb to obtain sur molecules/cell/min for the single BsAb (Fig. 65) to 660 mole- face binding, the temperature was then increased to 37Â°C,and cules/cell/min. Furthermore, this induced rate of uptake was the uptake of I25l-saporin was determined at various times. more than double that seen for the mixtures of BsAbs (300 Total cell-associated (surface plus internalized) and free 125I- molecules/cell/min). saporin were separated by spinning the cells through phthalate oils as described above. Internalized 125I-saporin was distin DISCUSSION guished from surface material by using papain to strip cell surface IgM along with any bound saporin/BsAb complexes. Previous work from this laboratory has shown that BsAb can As with the kinetics studies above (Fig. 5), to ensure that the be used to deliver saporin to guinea pig L2C lymphoma cells in uptake of saporin could be compared with different BsAb a highly selective manner both in vitro and in vivo (7, 8). The derivatives, irrespective of their binding avidity, concentrations immunological pressure exerted on the tumor in this way is of I25l-saporin were chosen that gave a constant number of such that antigen-positive lymphoma can be eradicated from saporin molecules bound to each cell. tumor-bearing animals, leaving only rare antigen-negative var The results in Fig. 6, A to F, show that, after 2-h equilibration iants to emerge eventually as new tumor. Throughout this work, at 4Â°C,all the derivatives, single and combined, resulted in BsAbs made with Fab' from polyclonal anti-saporin outper approximately 30,000 molecules of I25l-saporin bound/cell. formed those made with mouse monoclonal anti-saporin anti This level of surface binding did not show any major or sus bodies. In an effort to explain this observation and to construct tained changes during the course of the experiment. Any fluc more potent BsAbs from monoclonal antibodies, we have now tuations over the early time points were probably the result of prepared derivatives from an extended panel of five monoclonal new saporin-binding equilibria being established on warming to anti-saporin antibodies, with the specificity for the L2C cells 37Â°C.All the derivatives, single and mixed, resulted in signifi provided, as in the previous studies, by a monoclonal anti- cant accumulation of saporin within the cells over the 4-h time idiotypic antibody, anti-Id-1. 2358

antibodies when used as single reagents and gave equivalent protection to that provided by the original rabbit polyclonal anti-saporin derivative (8). The simultaneous use of two or more BsAbs constitutes a novel maneuver with potential bene fits for therapeutic applications. The superior performance of these mixtures and of the orig inal polyclonal reagent probably stems from the recognition of multiple epitopes on the saporin allowing multivalent attach ment to the target cells. Such attachment could have at least two consequences: (a) saporin would bind to the cell surface with an increased functional affinity; and (b) the internalization of saporin may be modified as a result of cross-linking of adjacent IgM molecules by the multivalent saporin/BsAb complexes. In agreement with the former interpretation, all our binding studies showed a positive relation between the functional avidity of the various derivatives and their ability to deliver saporin to the tumor cells. For example, the performance of the five single BsAbs correlated closely with the Kas of the monoclonal anti- saporin antibodies used in their construction (Tables 1 and 2). Thus, the three higher affinity antibodies, anti-sap-1, -3, and -5 (K, ~5 x IO7 M~'), produced derivatives that were superior to those made with the two lower affinity antibodies, anti-sap- 2 and -4 (Ka ~5 x IO6 M"')- Similarly, the combinations of BsAbs that performed so well in toxicity bound radiolabeled saporin to target cells 20 times more avidly than did the constituent single derivatives. The multivalency of these mix tures of BsAbs was indicated by the observation that, at satu ration, the number of saporin molecules bound/cell was only approximately half that obtained with a single BsAb (Fig. 4). This result is consistent with the expected 1:1 Bs,\Insaporili binding ratio with single BsAb and a 2:1 ratio when two BsAbs bind to saporin through different, nonblocking, epitopes. We have shown previously (9) that antibodies gain affinity by binding to cells through multiple Fab' arms. For example, the functional affinities of bivalent anti-idiotypic IgG and F(ab')2 were between 10 and 100 times higher than those of their univalent Fab' fragments (9). This work shows that a similar benefit applies to the multivalent binding of saporin to Fig. 6. Uptake of '"I-saporin by L2C cells in the presence of BsAb. BsAb (I target cells; a single BsAb, being univalent for both saporin and Mg/ml) and the required concentration of 12!I-saporin were incubated with L2C cells (5 x I0*/ml) for 2 h at 4'C. The temperature was then increased to 37Â°Cto idiotype, will not hold saporin at the cell surface following dissociation of either of its Fab' arms. However, saporin held allow uptake to occur, and samples were taken at the times shown. Intracellular radioactivity was determined by removing surface IgM and associated '"'I- by two or more BsAbs cannot escape unless at least two of the saporin/BsAb complexes by papain digestion, and total radioactivity, by sedi- four Fab' arms in the complex dissociate simultaneously. menting the cells through phthalate oils. Surface radioactivity was obtained by subtraction. The results are expressed as the number of molecules of saporin/cell. In comparison with its effect on affinity, our studies showed Total cell associated (â€¢);intracellular (O); surface (x). .1. anti-sap-1 x anti-Id-1 + 0.1 Mg/ml of saporin; B, anti-sap-3 x anti-Id-1 + 0.25 jig/ml of saporin; C, that the valency of saporin/BsAb complexes appeared to have anti-sap-5 x anti-Id-l + 0.25 Mg/ml of saporin; D, anti-sap-1 x ami Id-1 + anti- very little effect on saporin internalization. At least two mech sap-3 x anti-Id-1 + 0.025 fig/ml of saporin; E, anti-sap-1 x anti-Id-1 + anti-sap- anisms exist whereby surface molecules can enter the cell and 5 x ami Id I + 0.025 >ig/ml of saporin; and F, the cocktail of five BsAbs + 0.025 deliver antibody-bound RIPs: (a) constitutive turnover, during Mg/ml of saporin. In G, L2C cells were incubated with anti-sap-3 x anti-Id-1 and 0.25 ng/ml of saporin for 2 h at 4'C. Rat anti-guinea pig F(ab)/j (10 ng/ml) was which selected molecules are continually exchanged, often by then added to induce extensive modulation. The temperature was then increased recycling at the cell surface; and (b) ligand-induced uptake, to 37"C, and the time course continued as described above. when recycling is generally less prominent and internalized receptors are often degraded along with their ligands (22, 23). In the current work, in vitro toxicity studies showed again In the case of L2C cells, constitutive turnover allows IgM that none of these five monoclonal anti-saporin-containing molecules to persist on the surface with a half-life of about 5 h BsAb derivatives could match the activity of the original poly- (24), but it is not clear to what extent they are recycled back to clonal derivative. However, using a cocktail of all five deriva the cell surface, or whether they are simply degraded and tives, in effect a defined polyclonal reagent, or selected combi replaced by newly synthesized molecules. Whereas surface IgM nations of just two monoclonal BsAbs, we were able to equal, turnover is not perturbed significantly following interaction and sometimes outstrip, the performance of the polyclonal with a univalent ligand such as a Fab' fragment (25), cross- derivative. These results were confirmed in immunotherapy, linking with a bivalent antibody from polyclonal antisera results showing that a pair of monoclonal BsAbs, anti-sap-1 x anti-Id- in a dramatic and selective clearance of surface IgM molecules 1 and anti-sap-3 x anti-Id-1, outperformed either of these into the cell (21, 24), leaving the surface almost devoid of its 2359

Downloaded from cancerres.aacrjournals.org on September 23, 2021. © 1991 American Association for Cancer Research. BISPECIFIC ANTIBODIES FOR THE DELIVERY OF SAPORIN TO LYMPHOMA normal 100,000 to 150,000 molecules. In contrast, monoclonal relatively high binding constant (Table 1) produced a BsAb that antibodies cross-link much less extensively and consequently was 5 times less efficient than expected. This lack of activity induce a slower and less complete clearance of the surface IgM was underlined in binding studies which showed that, to achieve (9). an ICso with anti-sap-3 x anti-Id-1, 2 to 3 times more saporin We had anticipated that a single BsAb, binding univalently needed to be bound at the cell surface than with any other to surface IgM, would enter cells along with any attached derivative (Table 1). A possible explanation for this disparity is saporin through constitutive turnover. In contrast, a multivalent that the binding of anti-sap-3 interferes with the intracellular complex of saporin and two or more BsAbs might cross-link transport of saporin, possibly due to an unusually slow disso surface IgM in much the same way as a bivalent antibody and ciation rate of saporin/BsAb complexes or the blocking of an thus drive more saporin into the cell. However, direct measure epitope that facilitates the transfer of saporin to the cytosol ment of radiolabeled saporin entry into the cells showed no (22, 26). Alternatively, anti-sap-3 may bind an epitope that significant difference in the pattern of internalization induced enhances degradation of saporin within the cell, either by a by single or multiple BsAbs. It was particularly interesting that conformational change which opens the saporin to enzymatic during the accumulation of saporin in the presence of multiple attack, or by routing the complexes to the lysosomes where BsAbs, there was no concomitant loss of surface-bound saporin. degradation is likely (26). This indicated that the surface IgM was not being perturbed The results presented here highlight some of the advantages following binding of multivalent saporin/BsAb complexes and of using chemically engineered BsAbs as carriers of toxic com suggested that, as with the single BsAb derivatives, saporin was pounds. In particular, their precise construction has allowed being taken up by constitutive turnover rather than by the direct comparison of single and mixed reagents and has under induction of modulation. This interpretation is supported by lined the importance of functional affinity in targeting RIPs. the contrasting uptake pattern obtained following extensive, To date, our derivatives have been designed to deliver saporin independent, cross-linking of the surface IgM by a polyclonal through a single target antigen on tumor cells, but we are now anti-IgM antibody known to cause rapid internalization of in a position to undertake a careful assessment of different surface IgM. Under these conditions, the initial rate of saporin target antigens on human cells and, more importantly, the use uptake was increased (660 molecules/min/cell compared with of combinations of BsAbs which recognize different epitopes 300 molecules/min/cell obtained with a mixture of BsAbs), and on the saporin, or other RIP, and at the same time different there was a corresponding sharp drop in cell surface saporin. antigens on the target cells. In addition to maintaining high Because of the rapid clearance of surface IgM induced by the functional affinity for the RIP by multivalent binding, this polyclonal antisera, accumulation of saporin later in the incu approach has the potential to deliver more RIP to a cell and to bation period was effectively curtailed, and the total level of guard against the escape of variant tumor cells which have lost saporin associated with the cells (surface plus internal) over the the expression of a tumor-associated antigen (27). 4-h time course remained constant. This result points to a potential disadvantage of using antibody carriers that induce ACKNOWLEDGMENTS extensive internalization of their target molecule; while the initial rate of RIP uptake might be high, it may not be sustained, We thank Professor F. Stirpe, Dipartmento di patologia sperimen especially if the surface antigen is not replaced (21) and, at the tale, UniversitÃ di Bologna, for providing the saporin and for useful same time, the material accumulated in the cell is being de discussion. We would like to acknowledge colleagues in the Tenovus graded in lysosomes (23). Laboratory, particularly Dr. A. George for discussion of the manuscript. Thanks also to C. Pringle and S. Ranee for valuable technical assistance. The current results indicate that, for this system at least, the major factor influencing the performance of BsAbs is the avidity with which the saporin is bound to the cells, rather than the REFERENCES rate of internalization of saporin/BsAb complexes. The final 1. Vitella, E. S., Fulton, R. J., May, R. D., Till, M., and Uhr, J. W. Redesigning evidence to support this interpretation comes from a compari nature's poisons to create anti-tumor reagents. Science (Washington DC), son of the average number of saporin molecules needed on the 238: 1098-1104, 1987. 2. Blakey, D. C., and Thorpe, P. E. An overview of therapy with immunotoxins cells to achieve an IC50 with any of the individual or mixed containing ricin or its A-chain. Antibody 1nun unoci m. Radiopharm., 1: l - BsAb derivatives (Table 1). Binding studies showed that, with 16, 1988. the exception of anti-sap-3 x anti-Id-1, all derivatives delivered 3. Raso, V., and Griffin. T. Hybrid antibodies with dual specificity for the delivery of ricin to immunoglobulin bearing target cells. Cancer Res., 41: between 3,500 and 7,000 molecules of saporin to the cell at the 2073-2078, 1981. ICso concentration. At this level, between 2 and 5% of the 4. Raso, V. Antigen mediated delivery of toxic molecules to antigen bearing 100,000 to 150,000 surface IgM molecules present on L2C target cells. Imtnunol. Rev., 62: 93-117, 1982. 5. Corvalan, J. R. F., Smith, W., Gore, V. A., and Brandon, D. R. Specific in would be occupied by saporin. Similarly, the number of saporin vitro and in vivo drug localisation to tumour cells using a hybrid-hybrid molecules needed to achieve complete inhibition of protein monoclonal antibody recognising both carcinoembryonic antigen (CEA) and synthesis, again with the exception of anti-sap-3 x anti-Id-1, I'ini-u alkaloids. Cancer Immunol. Immunother., 24: 133-137, 1987. 6. Webb, K. S., Ware, J. L., Parks, S. F., Walther, P. J., and Paulson, D. F. was constant at about 20,000 to 30,000 molecules per cell. This Evidence for a novel hybrid immunotoxin recognizing ricin A-chain by one striking consistency in saporin levels gives a strong indication antigen-combining site and a prostate-restricted antigen by the remaining antigen-combining site: potential for immunotherapy. Cancer Treat. Rep., that it is the ability of different derivatives to capture saporin 69:663-672, 1985. onto the cell surface that determines overall performance in our 7. Glennie, M. J., McBride, H. M., Worth, A. T., and Stevenson, G. T. Preparation and performance of bispecific l-(;ili'->)? antibody containing assays. Once in the cell their potency is probably determined, thioether-linked Fab'y fragments. J. Immunol., 139: 2367-2375, 1987. as appears for conventional immunotoxins (23, 26), by the rate 8. Glennie, M. J., Brennand, D. M., Bryden, F., McBride, H. M., Stirpe, F., at which the RIP is translocated into the cytoplasm. Worth, A. T., and Stevenson, G. T. Bispecific F(ab'i)2 antibody for the The performance of one BsAb, anti-sap-3 x anti-sap-1, indi delivery of saporin in the treatment of lymphoma. J. Immunol., 141: 3662- 3670, 1988. cates that functional affinity for saporin is not the only factor 9. Elliot, T. J., Glennie, M. J., McBride, H. M., and Stevenson, G. T. Analysis that can influence toxicity. In this example, an antibody with a of the interaction of antibodies with immunoglobulin idiotype of neoplastic 2360

B lymphocytes: implications for imtnunotherapy. J. Ininnino!.. ÃŒ38:981- binding of multivalent immune complexes to Fc receptors. I. Equilibrium 988, 1987. binding. Biochemistry, 20:6326-6334, 1981. 10. Stirpe, F., Gasperi-Campani, G., Barbieri, L., Falasca, A., Abbondanza, A., 20. Peto, R. Editorial: Guidelines on the analysis of tumour rates and death rates and Stevens, W. A. Ribosome inactivating proteins from the seeds of Sapon in experimental animals. Br. J. Cancer, 29: 101-105, 1974. aria officinali* L. (soapwort), of Agrostemma githago L. (corn cockle) and of 21. Gordon, J., Anderson, V. A., and Stevenson, G. T. Loss of surface-bound Asparagus officinaux (asparagus), and from the latex of Hura crÃ©pitonsL. antibody accompanying the anti-complementary modulation of leukemic B (sandbox tree). Biochem. J., 216:617-625,1983. cell immunoglobulin: contrasting effects of antibodies directed against idi- 11. Shevach, R. M., Filman. L., Davie, J. M . and Green, I. L:( guinea-pig otypic and constant regions. J. Immunol.. 128: 2763-2769, 1982. lymphatic leukemia: a B-cell leukemia. Blood, 39: 1-12, 1972. 22. Hopkins, C. R. Membrane boundaries involved in the uptake and intracellular 12. Stevenson, G. T., Eady, R. P., Hough, D. W., Jurd, R. D., and Stevenson, processing of cell surface receptors. Trends Biochem. Sci., //: 473-477, F. K. Surface immunoglobulin of guinea pig leukaemic lymphocytes. Immu 1986. nology, 28: 807-820, 1975. 23. Youle, R. J., and Neville, D. M., Jr. Role of endocytosis and receptor 13. KÃ¶hler,G., and Milstein, C. Continuous cultures of fused cells secreting recycling in ligand-toxin and antibody toxin conjugate activity. In: C. W. antibody of predefined specificity. Nature (Lond.), 256: 495-497, 1965. Vogel (ed.), Immunoconjugates, pp. 153-169. New York: Oxford University 14. Fazekas de St. Groth, S., and Scheidegger, D. Production of monoclonal Press. 1987. antibodies: strategies and tactics. J. Immunol. Methods, 35: 1-21, 1980. 24. Glennie, M. J., Stevenson, F. K., Stevenson, G. T.. and Virji, M. Cross- 15. Stevenson, F. K., Morris, D., and Stevenson, G. T. Immunoglobulin produced linking of lymphocytic surface immunoglobulin inhibits its production via a by guinea-pig leukaemic B lymphoctyes: its source and use as a monitor of cyclic nucleotide mechanism. Nature (Lond.), 281: 305-307. 1979. tumour load. Immunol., 41: 313-321, 1980. 25. Gordon, J., and Stevenson, G. T. Antigenic modulation of lymphocytic 16. Markwell, M. A. K. A new solid state reagent to iodinate proteins: I. surface immunoglobulin yielding resistance to complement-mediated lysis. Conditions for the efficent labelling of antiserum. Anal. Biochem., 125:427- II. Relationship to redistribution of the antigen. Immunology, 42: 13-17, 432, 1982. 1981. 17. I-rignci, B., Charlotte, A. F., Djavadi-Ohaniance, L., and Goldberg. M. E. 26. Olsnes, S., Sandvig. K., Petersen, O. W., and van Deurs, B. Immunotoxinsâ€” Measurements of the true affinity constant in solution of antigen-antibody entry into cells and mechanism of action. Immunol. Today, 10: 291-295, complexes by enzyme-linked immunosorbent assay. J. Immunol. Methods, 1989. 77:305-319, 1985. 27. Glennie, M. J., McBride, H. M., Stirpe. F., Thorpe, P. E., Worth, A. T., and 18. Scatchard. G. The attractions of proteins for small molecules and ions. Ann. Stevenson, G. T. Emergence of immunoglobulin variants following treatment NY Acad. Sci., 51:660-672, 1949. of a B-cell leukemia with an immunotoxin composed of anti-idiotype and 19. Dower, S. K., De Lisi, C., Titus, J. A., and Segal, D. M. Mechanism of saporin. J. Exp. Med., 166:43-62, 1987.

2361

Downloaded from cancerres.aacrjournals.org on September 23, 2021. © 1991 American Association for Cancer Research. Cooperative Mixtures of Bispecific F(ab′)2 Antibodies for Delivering Saporin to Lymphoma in Vitro and in Vivo

Ruth R. French, Ann E. Courtenay, Susan Ingamells, et al.

Cancer Res 1991;51:2353-2361.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/51/9/2353

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/51/9/2353. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.